Ha ha, very funny, now read the rest of my post.
The assertion was that they didn't diversify the study groups enough. So long as they used gay people, this doesn't matter, as any "diversification" of the study group wouldn't add to the root factor: homosexuality.
An upper-middle class gay man with a home in suburbia, a nice car, and a supportive family and friend base is just as capable of displaying how homosexuality affects parenthood as a lower class gay man with a crappy job, living in the middle of a decaying post-industrial urban enviornment. The point is homosexuality, and any difference between the latter gay man and a straight man would probably have to do with his means - not necessarily his sexual orientation.