There are no imaginable behaviors that cannot have some outlandish, unprovable evolutionary justification attached to them. Just checking to see if I was still correct.If we actually cared about the nature of homosexuality, opposed to proving a political point, we'd check the incidence of homosexuality in social versus solitary animals. It's already known that pack animals are disproportionately homosexual(10% of male rams for example). If solitary animals contained fewer homosexuals it would go a long way towards proving the "beneficial to society" argument.
Or anything else that you wanted it to "prove", such as that an abundance of hormones in males makes them hyper-sexual, and that other hyper-sexual males are more receptive than the females.
In any event, your premise would provide no proof of societal benefit. For that you would actually have to observe a something called a benefit that could not be explained by other variables.