I wouldn't say narrow, only precise. I'm not denying group effects. Yes, bees are a good example. A bee does sacrifice itself for the hive, but the explanation that this brings an advantage to the hive is at best incomplete. The flaw of your theory is that it contrasts group advantage with individual advantage, both being irrelevant for natural selection. The bee sacrifices itself because saving itself would not help its genes.
The bee isn't capable of sexual reproduction, but even if it were, it would be maladaptive if it did. A bee is closer related to the other bees of the hive than it would to its own potential children! A bee sacrifices itself for only one reason: it increases the chance of the survival of the majority of its genes.

The same stringence must be applied to homosexuality. If homosexuality is adaptive, then it must increase chances for the genes. It has been hypothised that after several male children, then need to further offspring generating males diminishes. If only one of the heterosexual male children survives, he can create unlimited offspring. Therefore male children further down the birth order may better have more female traits like caring and socializing, without having children of his own because that may help to increase the chances of the offspring of the older males.
Having a homosexual child may indeed be beneficial because of social effects, but the interest is that of genes, not the group.