Summary: Since the original paper was published, more papers have come out with similar conclusions. It seems likely that individuals with higher pornography use are less likely to get married.
When I see an interesting scientific finding, I am always skeptical. Science works in fits and starts, and it can take decades before a new hypothesis becomes a proven (or disproven) fact. For an individual scientific finding, the best test of realness is what other peer researchers have to say about it.
These 11 scientific papers have cited the original. 4 other items at that link are either political or religious in nature, and I have discarded them. 1 other item is a duplicate. 4 scientific papers have a title that seems directly relevant to the question. Lets look at those first:
A pair of papers by S.L. Perry, looks at data and finds that greater pornography use leads to more breakups and fewer marriages. Both papers are longitudinal surveys; they follow a group of subjects over time and ask them questions about their porn use and relationship status among other things. This supports the claim.
This letter by M. Regnerus, is a positive response to this full article by Leonhardt et al. They argue that their evidence shows pornography is bad for long term sexual quality. They acknowledge that the literature has shown some positive effects of pornography, but argue that these positive effects are short term only. This also supports the claim.
This Ph.D. Thesis by A. Shaw looks at the effects of pornography on people in committed relationships. In the abstract Shaw writes that solo pornography use is associated with lower relationship satisfaction. This indirectly supports the claim.
Shaw also reviews a lot of scientific literature for and against the claim, inlcuding a criticism of the entire field of research on this topic.
In their review research linking pornography use to relationship functioning, Kohut and colleagues (2016) argued that a majority of research on pornography is harm-focused, seeking to demonstrate its adverse effects on individuals and their relationships: “Much of the empirical research concerning pornography’s impact on couple relationships can be legitimately characterized as a confirmatory search for the presumed harms of exposure” (p. 1). Kohut and colleagues (2016) went so far as to suggest that this negative bias is so pervasive and accepted in the field that few researchers have explicitly acknowledged this partiality. To directly address concerns that pornography research has overlooked possible neutral or positive effects, Campbell and Kohut (2017) recently used experimental methods to demonstrate benefits from pornography use (see below). Although that study represents an important first step, there remains a strong need in the field for unbiased studies of the effects of pornography use on individuals’ lives.
After some discussion, Shaw says "With the current study, I will take an unbiased approach, acknowledging that there may be negative effects of pornography use while at the same time examining potential neutral and positive impacts."
This Ph.D. dissertation references the study in the claiming paper, but describes it as using a "problematic method which relied on conflating respondents’ self-reported Internet use with self-reported pornography use..." This undermines, but does not refute the claim.
I have spent a couple hours skimming through these studies, which is not enough time to really comprehend them. Every single one of them references a dozen more studies, which would take hundreds of hours to read. This may be too big of a question for a non-expert like me to answer. Science may not have found a definitive answer to the claim, or the answer may not be as simple as presented in the claim.
From what I have seen, there is a substantial body of evidence generally supporting the claim that pornography use may discourage marriage or other long term relationships. As discussed above, there is some criticism of the entire body of evidence as biased, but I don't think it is enough to convince me to disregard the data collected by the above authors.
