Phil Arena has posted a very nice review of Clarke and Primo. Here is the money sentence:
There are of course differences between empirical models and theoretical models, but they are fewer and more modest than many empiricists realize.
I taught the PoP version of the book in Comparative Methods, but I bought the book myself to read, and I recommend it. It’s really helped me think through the enterprise of positive political science.
I can boil down the main lesson pretty well in a couple of sentences. Models are objects, not claims. As such, they do not have truth values (I should have been more careful on point 4 when writing this). The correspondence between theoretical model and an empirical model is complicated. That does not make models useless—there are many, many uses for empirical and theoretical models alike, and like it or not we use them anyway—but it does threaten the naïve version of hypothetical-deductivism that underlies a lot of what political scientists seem to actually be doing.