In fact, that has zero reporductive value.
That may not be true. Nicolaou linked the article about a 2004 study. http://www.timesonline.co.uk/tol/news/uk/article493668.ece
The findings suggest that the same genes that trigger homosexuality in men also promote fertility in women, and that this could explain how they survive in the population when gay men themselves are unlikely to breed. The genes are instead passed on through the female line and the enhanced fertility they confer on these women ensures that they are inherited by plenty of children.
Same group in 2008:
Male homosexuality is difficult to explain under Darwinian evolutionary models, because carriers of genes predisposing towards male homosexuality would be likely to reproduce less than average, suggesting that alleles influencing homosexuality should progressively disappear from a population. This changed when previous work by Camperio Ciani and collaborators, published in 2004, showed that females in the maternal line of male homosexuals were more fertile than average.
They concluded that the only possible model was that of sexually antagonistic selection. The other models did not fit the empirical data, either implying that the alleles would become extinct too easily or invade the population, or failing to describe the distribution patterns of male homosexuality and female fecundity observed in the families of homosexuals. Only the model of sexually antagonistic selection involving at least two genes – at least one of which must be on the X chromosome (inherited in males only through their mother) – accounted for all the known data.
http://www.physorg.com/news132983648.html