I have listened to all of the 'Medical Myths" lectures. They are very good, with an occasional error slipping in here and there. For example, ABO blood groups does accurately predict personality type. My blood type prediction was spot on (with the exception of being 'sociable'). Also ABO antigens aren't proteins; they're carbohydrates (that was probably a slip).
One thing that did have me worried; in the lecture on the placebo effect, Steve discussed 'confirmation bias' and I thought, he's going to talk about 'regression to the mean' next, and that's what he did. Has he smuggled in some subliminal psychic control mechanism into the audio. Obviously, the tin foil hat wasn't working.