I disagree with how far he takes his metaconclusions with which he judges his hypotheses. He rightly rejects bad scientific practices which throw away most data in a Manichean fashion and then use bad methods of analysis anyway to reach wrong conclusions but he then ends up with engaging in the massively motivated collection of confirming evidence that the scientific method is intended to prevent, the generation of cheap evolutionary just so stories, etc.
His theory of holidays, for instance, is a parody of bad evolutionary psychology. Evolution is all important, ultimately, but it's too dumb to make basic competence at hiding instinctual for toddlers. That's pretty bad!