I like that framing of the mutual interdependence of the two. Personally, I like beginning with pure theory, and then testing it against real world occurrence. I think in this way we get the best of logic and the best of empiricism.

The issue comes in the way we structure experiments though, because they rely on assumptions. Those assumptions really must be grounded in the soundest logic; otherwise, the entire endeavor becomes pointless.

As to the brain, I agree. Since this is the subject I spend the most time on. My attempts to explain human behavior have progressed as follows: using the logic of Darwinism (and evolutionary psychology) to draw up an explanation for most of what we do, and then trying to trace those behaviors back to specific brain areas, so pure theory backed by neuroscientific evidence.