Astronomy is mostly history sprinkled with physics.

You explained the limitations astronomers and medical researchers face. Psychologists face similar problems, which is why all their results should be treated with a certain amount of scepticism. But that does not mean their work is worthless; just that it is hard. A lot of traditional psychology was based on what one person thought, rather than logical arguments or experimental evidence. Evolutionary psychology is an attempt to place the study of the brain's workings in the context of evolution.

I've never heard anybody talk about how they expect behaviors to actually have formed over generations (nor does it meaningfully cover learned and taught behavior)

Individual human behaviours depend on a lot of other factors. All you can do from an evolutionary perspective is to explain some common trends. For example, in almost all cultures, some people are gay / ace. Traditional psychologists long thought of this as some sort of mental condition. But if you think of society in the context of inclusive fitness and r/K strategy, it makes a lot of sense to have a certain percentage of the population not reproduce. Is this why some people are gay / ace? I don't know, and I don't think we'll ever know. But at least we can try to explain some things.

Yes. It is unethical to give someone a disease so you can study it. Best we have are case studies of people who got the disease and are being treated for it.

In climate studies, it is not practical to increase temperature or humidity by x% and see the effects. Again, you have case studies - either from the past or from parts of the world that are warming much faster than the rest. Or you can do mesocosm experiments where you warm, say, a square metre of grassland, and see the effects. But then there is a lot of uncertainity in scaling up the findings of such small-scale studies.