In the case of ants, the beneficiary of the altruism/self-sacrifice is the queen, which always carries the gene. That's why I said:
>Biological altruism will generally mean an organism helping another at a small or no cost and/or the benefited organism is a close relative that likely shares the gene.
Ants and bees are extreme examples of the later.
In your theory, the only beneficiaries of fewer individuals reproducing are the straight (and bi) individuals who are still reproducing. It's a really bad theory to explain homosexuality and I don't know why anyone would insist on it when it's so easy to come up with better ones. Just think of something that would benefit the individuals carrying the genes themselves. Maybe an attraction to muscles (or some other manly characteristic in that species) that leads to stronger offspring when the individual mates with muscular females, but can go too far sometimes leading to homosexuality.
>one surprising (to me) thing that came up is that there have been substantial revisions in the estimated speed of evolution as a process since Darwin's days. Sexual selection in particular is a mechanism that turned out to work much faster than originally anticipated, and can in fact push other traits that wouldn't be viable otherwise over the viability threshold. One theory is that it developed in the first place because faster selection - i.e. adaptability - is itself a selected-for trait, so it's kinda getting very meta.
That's actually really interesting! Got any good links on that? The speed of evolution is something that sometimes bothers me when looking at highly complex adaptations.