Jane,
I have a hard time trusting the results of fMRI studies in general. Perhaps it's just because my field of study uses similar instruments as are used in medical imaging (indeed, many of them were invented for use in my field; it's one of our "killer apps"); we have enormous difficulties teasing out real effects from noise and bias in the analysis. FMRI studies seem to be equally susceptible to these difficulties, and for similar reasons.
Additionally, there is a huge question of interpretation. Any brain differences after a very young age might well be attributed to socialization; even in the case of trans people, it is possible that at a very young age, children raised in one gender identify with the other, and that this might provide a different kind of socialization.
So for me it's not so much a question of being politically correct, but scientifically -- and in my field, "scientifically correct" means a five sigma effect in a blind analysis, even with far fewer questions of causation.
But it doesn't matter. The existence of trans people means that gender differences are real and almost certainly innate. But those differences are ones that cut to the heart of the deepest questions of identity, perception, and consciousness. They aren't questions that lend themselves to simple answers. In the end, what matters is not root causation, but lived experience.
