No direct evidence, yet. For causation, and correlational, observational studies. All are too often based on animal studies. But these preliminary and circumstantial evidence point largely to smaller sizes being one quite plausible outcome.
Perfluorinated compound are a large group of man-made chemicals. While it seems quite necessary to look at research for each individual substance and its effects on human reproductive systems, one thing seems easy: since they are articial molecules (with very small number of unrelated to the claim exceptions), "higher levels" found in humans means "anything above 0".
The main classes of molecules to evaluate here are Perfluorooctanoic acid (PFOA), Perfluorooctanesulfonic acid (PFOS) and closely related substances. Like PFBS, which is also persistent in the environment and not adequately tested for their risk-level for humans. They are so well known as bio-persistent and bio-magnificating, environmental pollutants that lawyers and governments have successfully forced manufacturers to reduce or stop their production in their 'traditional' production sites. (EPA: Emerging Contaminants Perfluorooctane Sulfonate (PFOS) and Perfluorooctanoic Acid (PFOA), 2014, unstable-link). Meaning: Production in Europe and North America was shifted to Shandong in China.
But not only production is hazardous, kitchen ware, outdoor clothing, fire extinguisher foams and a huge range of other applications, together with all of those when they go to waste, ensure that these substances get world-wide circulation. Note that pure PTFE/Teflon where the aforementioned substances are supposedly only used in manufacturing, is less hazardous, but except for medical applications in humans, "pure PTFE" is a rare find on the market, as often the much more problematic PFCs are not 'cleaned out' properly.
Illustrating the difficulties already encountered in rats: there are large differences between the different chemicals studied, and large differences when comparing one and the same substance between male and female rats.
We evaluated pairs of studies performed with different perfluoroalkane acids in the same species using the same design and found that endpoints for perfluorooctanesulfonate (PFOS), perfluorooctanoic acid (PFOA), perfluorobutanesulfonate (PFBS), and perfluorodecanoic acid (PFDA) could be discordant. We evaluated pairs of rat studies of PFOS, PFOA, and PFBS performed with the same design for which dose–response curves could be modeled for the concordant endpoints, but we were unable to identify a scaling system that gave values consistently within an order of magnitude for the same compounds. Currently available data do not support the combining of exposure levels of perfluoroalkane acids for risk assessment, although re-evaluation after additional data are available is recommended.
Administered dose was used in all cases; in addition, associated serum or plasma concentrations were considered when available. PFOA undergoes rapid renal elimination in female but not male rats, probably due to sexually dimorphic organic anion transporter expression (Kudo et al., 2002). The excretion of PFOA by female rats is virtually complete within a day. It has been recommended that PFOA serum concentrations be approximated in female rats by dividing the area under the time–concen- tration curve by 24 to give a time-weighted serum concentration over the course of a daily (24 h) dosing period (Butenhoff et al., 2004a).
Anthony R.Sciallia et al.: "Combining perfluoroalkane acid exposure levels for risk assessment", Regulatory Toxicology and Pharmacology
Volume 49, Issue 3, December 2007, Pages 195-202. DOI
Moreover: it seems that these effects are more studied in their effects on females than males.
"These findings suggest that PFOA and PFOS exposure at plasma levels seen in the general population may reduce fecundity; such exposure levels are common in developed countries. (DOI)"
In adjusted regression analyses, daughters exposed to higher levels of PFOA in utero had a 5.3 (95% confidence interval: 1.3; 9.3) months later age of menarche compared with the reference group of lower PFOA. Crude (P = 0.05) and adjusted (P = 0.01) trend tests also indicated a relationship between higher prenatal PFOA exposure and delay of menarche. (DOI)
The most robust finding in the present study was the negative associations between PFOS exposure and sperm morphology suggesting adverse effects of PFOS on semen quality, possibly due to interference with the endocrine activity or sperm membrane function. (DOI)
PFOS levels were significantly negatively associated with serum testosterone (total and calculated free), but not with any other reproductive hormones or semen quality.(DOI)
For example, the perfluorinated toxicant perfluorooctane sulfonate (PFOS) has been shown to significantly reduce circulating thyroid hormone levels in mammals. (p146)
Perfluorooctanoic Acid (PFOA)
Perfluorooctanoic acid (PFOA) is a chemical used in fire-fighting foams, electronics, and making commercial products grease and waterproof, and has been found to be the final degradation product of other>8-carbon perfluorinated materials. Persistent traces of the compound have been found in humans and wildlife alike, making it a prime target for toxicity studies, especially developmental effects . Various studies on the mammary gland rodent model have indeed found altered development after gestational PFOA exposure.
Research has shown significant delays in gland development after the gestational exposure of mice to PFOA, both in dams and offspring. A dose of 5 mg PFOA/kg/ day administered to CD-1 mouse dams during gestation altered the lactational ability of dams. The dams dosed during GD 8–17 and 1–17 displayed severely delayed mammary gland development on PND 10, the peak of lactation. This dose administered from GD 12 to 17, GD 8 to 17, or GD 1 to 17 caused significant retardation of mammary gland development, proliferation, and differentiation in the offspring. Pups also had decreased body weight compared to their vehicle-treated counterparts, but it was not found to be a significant variable in their mammary growth disparity .
Other studies utilizing cross-fostering or restricted gestational exposure designs reported that gestational and/or residual lactational (in milk) exposure to PFOA caused detrimental effects to the developing mammary gland. In control mice offspring cross-fostered to dams exposed on GD 1–17, developmental deficits were apparent from the initial sacrifice day of PND 21 and up to 9 weeks of age. These lactationally exposed pups had low serum PFOA levels comparable in magnitude to levels found in humans exposed to PFOA through contaminated drinking water or occupational hazards, making these findings much more human relevant. Similar mammary gland developmental delays were observed in the exposed offspring cross-fostered to control dams as well as those receiving gestational and lactational exposure. The study also found that developmental effects of gestational-only and/or lactational-only exposed mice could be observed as early as 12 h after parturition _ENREF_65 [71, 72]. These immediate effects are proposed to be a result of an interruption of the rapid growth and development of the mammary gland parenchyma that takes place in the first day of postnatal development 7. Recent studies [73, 74] define even lower levels of PFOA (5 ppb in water; 0.01 mg/kg/d from GD 10 to 17, respectively) as effective in delaying mammary gland development, and the internal doses overlap with those reported in humans living in PFOA-contaminated communities in OH and WV, USA .
Young adult PFOA exposure also altered mammary gland development. In BALB/c mice, a dose of 5 or 10 mg PFOA/kg administered for 4 weeks, starting after weaning, caused reduced ductal length, decreased number of terminal end buds, and decreased stimulated terminal ducts in the mammary glands. The number of proliferating cells in the ducts and terminal end buds were significantly lower than those in controls. In C57BL/6 mice, the same doses caused a different response. At 5 mg/kg, a stimulatory effect was observed with a significant increase in the number of terminal end buds and stimulated terminal ducts. At 10 mg/kg, mammary growth was delayed . Further examination of the mechanism behind the stimulation observed in C57BL/6 mice revealed that PFOA may alter hormones, growth factors, and cellular receptors that promote mammary gland cell proliferation .
This is a host of first line evidence in rodents. However, applicability to humans is of course limited, as not enough studies investigated these possible links.
Many toxicology studies on rodents today address pubertal end points such as first estrous, estrous cyclicity, and vaginal opening. Although these are important markers in determining rodent puberty, they do not directly apply to human pubertal mechanisms and end points. Including mammary gland development as an end point in these studies creates a vital bridge between animal studies and human relevancy. But things do not look good or assuring.
As sections of this chapter have pointed out, EDCs have been shown to affect male mammary gland development as well. These findings shed light on a major gap in mammary gland development research. Including male development in studies and determining which environmental compounds affect this sensitive end point should be a key goal as mammary gland development research continues to discover and identify compounds negatively affecting this tissue.
Suzanne E. Fenton & Lydia M. Beck et al.: "Developmental Exposure to Environmental Endocrine Disruptors and Adverse Effects on Mammary Gland Development", p 201–224, in: Evanthia Diamanti-Kandarakis & Andrea C. Gore (Eds): "Endocrine Disruptors and Puberty", Humana Press: New York, Dordrecht, 2012.
The focus on males and consequently their reproductive systems is similarly well studied, or rather less well. The amount of possible but uncontrolled for confounders is massive, but with preliminary research raising concerns:
Polyfluoroalkyl chemicals (PFCs) are widely used in commercial and industrial applications as surfactants, paper and textile coatings, and in food packaging due to their water and oil repelling properties . Biomonitoring studies have documented widespread human exposure to PFCs ; major sources of human exposure are not definitive, but may include diet (either directly from food or migration from food packaging), drinking water, and house dust (reviewed in ). PFCs bind to PPAR- alpha, a receptor closely involved in lipid metabolism, as well as to PPAR-gamma . In mice, prenatal exposure to a low dose of PFOA resulted in increased leptin and insulin levels at mid-life . There have been no prospective cohort studies to evaluate pregnancy levels of PFCs and effects on obesity in offspring, and findings have been mixed from cross-sectional studies assessing the relationship between PFCs and body weight in adults [154–155].
Four studies have suggested that PFCs may reduce fetal growth, although results have been somewhat inconsistent. In a study of 1,400 women and their infants randomly selected from the Danish National Birth Cohort, head circumference, abdominal circumference, birth length, and placental weight were all reduced in infants who had been exposed to higher levels of perfluorooctanoic acid (PFOA), one of the most prevalent types of PFCs. A cross-sectional study of 239 births evaluated cord serum levels of PFCs in relation to birth size and gestational length . Both perfluorooctane sulfonate (PFOS) and PFOA were associated with decrements in birth weight, ponderal index, and head circumference but not with gestational length after adjustment for lipids and other potential confounders. Two other studies suggested that PFOS, but not PFOA, was associated with reduced birth weight [158, 159].
To our knowledge, only one study has investigated the association between prenatal exposure to PFCs and postnatal growth . Using data from the Danish National Birth Cohort, Andersen et al. found that infants with higher prenatal exposure to PFOS and PFOA tended to have lower weights and body mass index at 5 and 12 months of age, after adjusting for birth weight. When the data were stratified by gender, effects were much more apparent among boys and there was little evidence for an association in girls. (p306)
Elizabeth E. Hatch & Jessica W. Nelson, et al.: "Developmental Exposure to Endocrine Disrupting Chemicals: Is There a Connection with Birth
and Childhood Weights?", p283–324, in: Evanthia Diamanti-Kandarakis & Andrea C. Gore (Eds): "Endocrine Disruptors and Puberty", Humana Press: New York, Dordrecht, 2012.
Since the focus of the claim is "penis size", which is primarily growing during puberty:
Perfluorochemicals Perfluorooctanoic Acid (PFOA) and Perfluorooctane Sulfonate (PFOS)
Members of the perfluorochemical (PFCs) chemical family in polymeric form have applications in waterproofing, protective coatings, nonstick coatings for cookware, electric wire insulation polishes, flame retardants, food packaging, and other uses. Due to their widespread use, chemical stability, persistence in the environment, and bioaccumulation in humans, animals and wildlife, PFCs have been a subject of concern to scientists and citizens alike [71–73]. In 2000, the primary global manufacturer began to phase out a major PFC after a reported end product perfluorooctane sulfonate (PFOS) was found to bioaccumulate in humans and wildlife [74, 75]. In the subsequent 5-year period following phase out, there was a decline in blood plasma concentration of PFOS in specimens obtained from US regional blood banks . The reported geometric mean concentration (PFOS) post phase out was 14.5 ng/ml. These data differ somewhat from serum concentrations reported  in a population-based national sample (geometric mean 20.7 ug/l) obtained in the third NHANES study survey (years 2003–2004). It appears that continuing sources of human exposure to PFOS remain.
PFOA and PFOS are commonly detected in human serum, more so in the US population than in the developing countries of Asia and South America . The overall half-life of PFOA in human serum is about 3.5 years and for PFOS about 4.5 years. Concentrations in males tend to be higher than in females . Curiously, since PFOA and PFOS are both lipo- and hydrophobic, rather than bioaccumulating in fat, PFOA and PFOS bind to serum protein .
In regard to puberty and puberty onset, there are a number of epidemiologic studies each having a different approach and outcome. As part of a breast cancer study , a correlation between girls arriving at Tanner stage IIB (breast development) and the concentration of PFOA (median concentration detected 6.4 ng/ml) was identified. In a cohort study of British mothers and their daughters, PFCs in serum samples obtained in pregnancy were correlated with age of onset of menarche of their female offspring. In this report, the authors state no correlation of PFCs with onset of the daughters’ menarche. However, authors mention that serum concentrations of carboxylates (including PFOA and other perfluorinated carboxylates) were associated with increased odds of earlier menarche . The median concentration of PFOS (19.8 ng/ml) and PFOA (3.7 ng/ml) were in the same range as the compiled results  from NHANES (I thru III) studies dealing with adult women (weighted mean of PFOA is 3.77 ng/ml and PFOS is 19.14 ug/ml).
A cross-sectional study  was done on children who were residents of a community living near a fluoropolymer manufacturing plant where drinking water was contaminated with PFOA. Following parental consent, serum specimens were obtained from children ages 8–18 years. Measurements of PFOA and PFOS were conducted and correlated with sex hormone measurements. Self-reported questionnaire data was used to assess age at menarche. No comparable questionnaire information was obtained for boys. Total testosterone level greater than 50 ug/dl was set as the cutoff point to determine onset of puberty in boys. The median PFOA and PFOS serum concentration in the children was 28.2 and 20.0 ng/ml, respectively. The median serum concentrations of PFOA (4.2 ng/ml) and PFOS (17.5 ng/ml) determined in NHANES survey collected in the same time frame (2005–2006) for PFOA may be significantly different.
For boys, increasing PFOS was associated with delay of 190 days in the onset of puberty. For girls, higher concentrations of PFOA or PFOS were associated with reduced odds of entry into menarche (130 and 138 days delay, respectively). Several animal studies suggest that higher PFOA concentrations are associated with lower testosterone levels in male rats  and higher progesterone levels in female mice . While initial animal studies suggested that activation of peroxisome proliferation might play a major role in the endocrine disrupting effects of PFOA, more recent work point to specific effects on steroid metabolism . Another work points to disruption of thyroid hormone action . (p366/7)
Vincent F. Garry & Peter Truran: "Secular Trends in Pubertal Timing: A Role
for Environmental Chemical Exposure?", p357–372, in: Evanthia Diamanti-Kandarakis & Andrea C. Gore (Eds): "Endocrine Disruptors and Puberty", Humana Press: New York, Dordrecht, 2012. For the book and all the further references, DOI: 10.1007/978-1-60761-561-3