I think his thesis is plausible and the supporting comments in the text and the copious footlights seem to imply there is proof, but I would have liked to see more statistics or demographic maps to be convinced. Perhaps even demographic, electoral or statistical maps and data throughout history to show that these regions really are different and have remained coherent over time.
I also imagine that even within the different nations there are further divisions due to race or socio-economic differences. While I agree that the white leaders of the Deep South have one agenda, the African American populace in that region has a different one and their growing political power will have implications. I could use Philiadelphia and it's environs, the birthplace of the Midlands, as an example of how demographics and socioeconomic status can have an effect within regions. Philadelphia often votes very differently from its surrounding counties, seemingly making It hard to use the fact that it is in the Midlands to predict an electoral or political outcome.
I recommend the book for its coverage of history, some of it forgotten (several rebellions before the revolutionary war) through a different lense than more traditional histories. History may well prove Colin Woodard's thesis and predictions correct in the long run.