I suspect that there is something of a simple rationale for homosexuality in basic biology. That is, all people are composed of both dominant and recessive genes. Generally the dominant genes are better, and the recessive genes are “backups”, sometimes defective, but available if the dominant genes “fail” in some way, over a single, or several generations.
This being said, since humans have relatively few offspring, another genetic methodology would be to “load” the first child with a predominance of dominant genes, and give subsequent children increasing numbers of recessive genes, so that they continue in the bloodline.
However, beyond a certain number of recessive genes, and a child may not be viable for reproduction. For this reason, a “safety” is engaged of making them homosexual, since any offspring they had would be genetically worthless.
Importantly, the first theory seems to hold water, because it has been statistically noted.
“In several studies, the observation is that the more older brothers a man has, the greater the probability is that he will have a homosexual orientation. It has sometimes been called the ‘older brother effect’.”
While I haven’t seen any studies about predominance of recessive genes among homosexuals, it is widely known that as a group, homosexuals have far more physical and psychological problems *not* associated with their sexuality, than average. This would seem to support a recessive “loading” “safety” theory.
Importantly, homosexuality should be looked at as just one mechanism by which reproduction is prevented. Non-fertile prostitution, and sexuality in post-menopausal females might be two other ways in which those who should not reproduce are distracted away from those who should.
No, dominant genes are not “better”. Dominant genes have a stronger effect (for good or bad) than recessive genes. This usually has to do with the molecular biology of the resulting protein from the two different alleles.
If wrinkled pods are dominant to smooth pods in a plant - it isn’t because wrinkled pods are better - it is because the allele that specifies wrinkles is sufficient in just one copy to determine that the pods are not smooth.