Well, on that, we agree. The question becomes, is it even theoretically possible to arrive at a "final answer", or are we merely pawing at reality with our formulae like a teenager trying to undo a bra?
This largely depends on what it means for scientific theories to have "final answers," in your view. In the meantime, I'll assume you're suggesting that a scientific theory must admit of exceptionless regularity (i.e., no anomalies or counterinstances) in order for it to be considered "final." In theory, I think it is possible to gain knowledge of this sort. For example, Newtonian Physics was in some respects superseded by Einsteinian Relativity, but the advances made in Relativity were only possible because they retained precise Newtonian predictions and approximations, which still hold at the macroscopic level.
Should this process of retention and refinement between theories persist, then perhaps, with time, a final answer will be given. If our scientific theories do give approximately true descriptions of a mind-independent reality, as I believe they do, then perhaps science is
closing the epistemic gap between our knowledge and the final answers. Then again, perhaps we already do have the final answer in some places, partly because it is not a necessary condition of a physical law that it admit of exceptionless regularity. In other words, many physical laws are ceteris-paribus generalizations. Consider the following statement: "smoking causes cancer." Phrased as an exceptionless regularity, this statement cannot be true, because we have counterinstances of smokers who have lived and died without ever contracting cancer. But if we say instead "ceteris-paribus, smoking causes cancer," this generalization is true, if incomplete. Perhaps this isn't too convincing. But either way, I think we have good reason to be optimistic about getting physical laws right (assuming I have correctly assumed what you mean by "final answer"). Lastly, it is worth mentioning that if we insist on certainty
as a precondition for knowledge, then we can only know the most trivial truths (like the momentary contents of one's thoughts).
Next question: if we live in a non-deterministic universe (that is, if free will exists), then how can a predictive physical or moral model could ever have it "exactly right"? Is free will not, perhaps, an x factor that torpedoes the whole project?
Your underlying assumption here is that free-will is only compatible with an indeterministic universe. Compatibilist philosophers believe you can reconcile free-will (properly understood) with determinism. One such view is defended in Daniel Dennett's Freedom Evolves.
one restricts the definition of free-will to its libertarian locution, i.e., "the ability to do otherwise," and if
one insists that only an indeterministic universe is compatible with libertarian free-will, then whether you could have an accurate predictive model depends on what you believe is really at stake if indeterminism is true.
But consider: does free-will require freedom from causality itself? Compatibilists answer in the negative. At any rate, it's definitely food for thought.