If you don't like reduced form graphs of change measures, you're going to have a really hard time with every Kenworthy book. I believe KW mentioned that the conclusions are also drawn from FE models.
I _think_ the viz. decision was to allow for a broader readership.