I agree that it is complicated, but if this kind of analytical framework reduces the party to a weathervane that only plays it safe (yes, this is an oversimplification), what are we left with? Couple this with the fact that we can't know the future, as political society is a complicated system, I would like to avoid this type of thinking as much as possible.
As to the Weimar example - we are acting with the benefit of hindsight and one still can't say "But for the Weimar decision on the treatment of Jewish folks, Hitler would not have come to power." I just think that kind of analysis can lead to bad outcomes.
I think there are two issues here.
1. Your second paragraph is pointing to an empirical consideration -- namely, that the future is unknowable. And you're definitely not wrong here. Going back to that Weimar example, I absolutely could see myself, if put in that position, thinking something along the lines of: "I don't know what will happen in 10 years if I do this. It's out of my control; it will depend enormously on the actions of other people, most of whom I don't know; and even if it knowable in some abstract sense, it's not known to me. I have to do the best thing in my little sphere of influence, and that would be to promote integration."
And this is part of the reason why I characterize my earlier "fear" as a "fear," and the main reason why I don't ever put this point strongly. It's really hard to analyze things counterfactually. Even something seemingly simple, like "if only Pat Sullivan had hit that FT, we wouldn't have needed Chris Webber's fuckup" isn't really correct. We still could have lost that game, in regulation even (as Antonio Davis and the rest of the Pacers know all too well). And so it's really hard to say with confidence that if the liberals hadn't jumped to the defense of trans, the election would be better for us right now. But it's also hard to say that it wouldn't be.
So on the "future is unknowable" point, we're in agreement and it's surely important.
2. On the other hand, we would be irresponsible policy makers if we didn't at least try to predict the future. I mean, that's really what policy is all about. If we pass X law, will Y or Z be the result? Y is good; Z is bad; and the decision often comes down to a prediction of whether Y or Z is better.
Moreover, if the future is unknowable, then perhaps that means we should be considerably less risky in our policy choices. If we don't know whether advocating for trans people will elect Trump or not, then we have to start thinking about magnitude of harms. On one hand, if we do nothing, trans people will be subject to the same sort of discrimination that has existed pretty much everywhere Abrahamic religions have been strong. That's a harm, but let's compare it to the harm of a Trump administration. I would say Trump being elected would be almost infinitely worse. A Trump election means, among other things, that the world will pretty much just give up on fighting climate change. Our only hope will be technologies that remove carbon from the air and water. And that is a far worse outcome than trans kids being depressed. As someone who has battled mental illness all my life, I'm not saying that disparagingly. It's just that depression is manageable; the loss of 1/3 of our arable land is really not.
So I get where you're coming from about the weathervane, but in a democracy that's an essential aspect of politics. Always trying to do the right thing is a great way to make sure that the wrong things always happen. Bill Clinton personally oversaw the execution of Ricky Ray Rector. Perhaps it would have been better, in isolation, for him to grant Rector clemency. But if that decision was integral to him winning the presidency, then the loss of that one life was absolutely the right thing to do. We would have lost Roe for sure, and we'd probably have had back then the same lunatic court we have now.
This is why politics is not a game for the idealistic or the weak of heart.