The original study was by a group of researchers lead by Robert Zajonc. (Wikipedia attempts to describe the research, but does a rather ordinary job in this case.)
Convergence in the physical appearance of spouses
R. B. Zajonc, Pamela K. Adelmann, Sheila T. Murphy and Paula M. Niedenthal
MOTIVATION AND EMOTION Volume 11, Number 4, 335-346, (1987) DOI: 10.1007/BF00992848
This study attempted to determine whether people who live with each other for a long period of time grow physically similar in their facial features. Photographs of couples when they were first married and 25 years later were judged for physical similarity and for the likelihood that they were married. The results showed that there is indeed an increase in apparent similarity after 25 years of cohabitation.
They also proposed some mechanisms for this, which I didn't find awfully convincing (and neither did some other scientists quoted in your original article.)
Their results is contradicted by :
Shared experiences and the similarity of personalities: A longitudinal study of married couples. Caspi, Avshalom; Herbener, Ellen S.; Ozer, Daniel J.
Journal of Personality and Social Psychology, Vol 62(2), Feb 1992, 281-291. doi: 10.1037/0022-3522.214.171.1241
Do spouses become more similar over time? What processes contribute to enduring similarities between them? Using the 20-yr Kelly Longitudinal Study of couples, no support for the hypothesis that couples increasingly resemble each other with time was found. Rather, couples maintain the same degree of similarity across 20 yrs.
I've only found two contradictory studies to choose between, so it is hard to form a firm conclusion, but my interpretation would be that the Zajonc paper was a preliminary investigation in 1987 that suggested there may be a correlation worth following up, and the Caspi paper in 1992 finding that any effect was just an anomaly.