"> Is coffee good for you? What about wine or chocolate? Scientists trying to answer these questions"

These are dumb questions to ask in the first place, because the "you" and "good" here are too personal for any general answer to be useful to most people. Unfortunately, this is not just lazy writing that took complex questions and simplified them to the point of uselessness - we really are asking these kinds of questions :(

Most of this doesn't generalize to populations the size of the world in the way something like "physics" does, because, for starters, we aren't very deterministic or very homogeneous at large scale.

Instead, you end up with millions to tens of millions of people in a subgroup particularly affected or unaffected by something because of genetic variation, etc.

Any reasonable scientist knows this. Instead, the main reason to try to answer these questions framed like this seems to be either to get funding, or to make headlines.

Sometimes we can answer extreme versions of this question (IE it seems data suggests alcohol is fairly universally bad for almost any person, definition of bad, and amount), but that's pretty rare. This then gets used as a "success" to do more poorly designed and thought out studies.

Just because we want to know things doesn't mean we should use mechanisms that we know don't work and produce mostly useless results. This is true even when we don't have lots of mechanisms that do work or produce useful results.

It's much slower and much more expensive, but what we learn is at least more useful.

It's really hard, slow, and expensive to answer questions about particle physics - this doesn't mean we revert to asking atoms to self-report their energy levels and publishing headlines about how "larger atoms that move around more live longer" or whatever based on the results. Instead, we accept that it will hard, slow, and expensive, and therefore, we better get started if we want to ever get somewhere.