the result of a process complex enough that it's very difficult to predict the outcome or identify the root cause.

Difficult's a two-place word, and so I'm not sure it makes much sense to argue about whether or not something is 'objectively' difficult, instead of difficult at various states of knowledge.

it carries with it an assumption that disowning gay sons has evolutionary roots, or is ingrained behavior in humans, or is common.

It's not quite that disowning gay sons has evolutionary roots, but that disowning gay sons is not so heavily disfavored as to be extincted. For example, cultures where childbirth is prohibited mostly die out, and so on. But even less obvious things that have an effect on reproductive success are strongly motivating; in cultures with prohibitions against masturbation, those prohibitions are mostly not followed; in cultures where doctors tell mothers to avoid touching their infants because of disease risk, those prohibitions are mostly not followed, and so on. (The impulse for mothers to touch their babies seems very strong, and also very healthy--it actually lowers disease risk by informing the mother what antibodies she needs to produce for her child, and seems critical for proper psychological development.)

And traditional behavior gives us an imperfect window into the economics of the past, which is what's under discussion when we talk about historical selective fitness. If gay sons were helpful enough with nephews and nieces that it was as if they had had their own children, it seems to me they would be welcomed and lauded as examples of loving selflessness. But if gay sons were reproductively disadvantageous, and in particular if it was reasonable to expect that homosexuality is contagious, then there's little cost and some reproductive benefit to forcing them out of the home.

(I should note that the hypothesis that one gene causes both female fecundity and male homosexuality is also consistent with disowning gay sons, but I think that one has other challenges.)

After all, my family didn't disown me, and it seems like disowning gay sons is becoming increasingly uncommon.

Thankfully, people are much more motivated today by individual and relationship satisfaction, neither of which disowning is helpful with. (My family didn't disown me either.)

Does it make sense for families to disown children for being the wrong religion?

It suspect it made sense for religions to disown members that fail to disown their children for being the wrong religion.

Does it make sense for families to disown children for being pregnant?

It looks like control over sexuality was a big deal, and as a first-order effect it seems that signals of that control would heavily impact someone's price on the sexual marketplace. As a second-order effect, it seems that the harsher penalties are for not being controlled, the more likely people are to submit to control. But in less status-stratified societies? Probably not--and it seems like this is mostly a class thing in the societies that I'm familiar with.

This is mostly my speculation, though--I haven't read much on evolutionary accounts of how parents should respond to teenage pregnancy in various environments. I expect someone has thought about this problem.