The only "microscopic" theory of superconductivity that is widely accepted in the BCS theory. You can easily look at its derivation from First Principles from the paper itself and satisfy yourself on its many-body starting point.
Free will is a topic with a lot of meat on it and there are many ways I'd like to slice it up. However I'll try to restrict this to aspects that are more relevant to this post.
There is a philosophical definition of truth which I've often heard scientists use that I think applies here. Usually the way it is stated is different but my phrasing is that "A proposition contains truth within a given context, to the extent that it is useful." I've often heard that a given theory or hypothesis is "good" because it makes useful predictions for example. There are many ideas that have been "true" in this sense historicaly, that have later been pushed aside to one degree or another when a "better" theory came along.
For example Newtonian gravity was "replaced" by relativity. Einstein's version is considered to be more useful, and thus more true or accurate. The same flow is apparent in the historical progression that led to QM. Is QM "complete"? It's difficult to say (my personal hope is that it is but then I'd like a million dollars too and I'm not willing to go through the work required to get that either.) What is not really debatable is that QM is useful.
However, trying to predict the outcome of human behavior with either classical or quantum mechanics isn't going to get you very far. This doesn't imply that the mind is non-physical per say, but rather that mechanics isn't the most useful set of principles to use in that realm.
I feel that Free will is a useful "truth", in that it apears to explain human behavior more accuratly than other concepts.
However free will is certainly not facilitated by randomness (I think the QM -> FW crowd is inserting a soul in to the cracks without realizing it) Free will implies that a person makes decisions based on rules, and evidence. In this sense Free will is a deterministic theory anyway.
Consider that a perfect intellect with access to perfect information will always make the same decision in the same circumstance. This does not imply that the entity has no choice, rather it is the result of making a "good" decision. While free will implies that the entity could make another choice, the less "good" choice wouldn't make sense. That is that one could decide to light themselves on fire just to prove that they can make such a choice.
This also implies that a perfect Free Entity will behave (to the extent that it is able and interested) in a predictable fashion more or less. However humans while they may have free will, do not have perfect knowledge. Still most human behavior ends up withen fairly narow bounds. For a given situation, most folks will behave in one of a finite set of ways.
It is I think in principle sensible to say that one could develop a theory that would statistically predict the outcome of a human interaction, if the initial conditions were known in detail. Such a theory would not be able to predict a specific person's response however as the Human's internal processes are hidden from us. Rather you could say something like "a given human in situation X has a probability distribution of Y"
I think it's interesting that QM has exactly the same problem ;) And for basically the same reason.