I haven't had time to read all of this thread, but 'homosexual' type behavior is prevalent in many animal species and clearly has all kinds of fitness-positive effects. However, most of this is 'facultative' homosexual behavior that either serves to reduce aggression, or to (indirectly) increase chances of mating with females.
I just want to point out that only 'exclusive' homosexuality would be selected against by natural selection and that is the exception rather than the rule. The only way to demonstrate NS advantages for exclusive homsexuality would be through models of inclusive fitness.