His position is a little more nuanced than that, but in general I disagree with his “either genes or parasites” approach to heritable phenotypes. I know a pair of identical twins where only one is gay, and it was obvious at least in middle school. I even remember associating the vowel at the end of his name with his softer features and voice, and the consonant at the end of his twin’s name with his more masc demeanor. I didn’t really even know what gay was in middle school. I just used that to tell them apart, and now that one of them is a male model in NYC and the other one has a wife and kids my old trick makes sense.
I suppose Cochran would say that, between birth and middle school, one twin was exposed to communicable gay and the other wasn’t. I don’t buy it. I think it’s a differential fetal testosterone exposure thing (same reason identical twins can sometimes be discordant for autism, despite autism being highly heritable) which means it’s probably adaptive in some obtuse way that Cochran would harrumph about (e.g. group selection, eusocial stuff, lower mate competition between siblings, more likely (in the past) to stay home and take care of dear mother, etc.).
Cochran is correct to point out that a dramatic change in effective reproductive phenotype should not persist at this fraction without either being advantageous or environmental.