You raise some excellent points, Witt, but I've got to disagree with you on a few. Disregarding or invalidating a person's emotions and their phenomenological reality is not a tenant of CBT. Providing accurate empathy and a solid understanding of the client's world isn't stressed in the literature for the same reason that "having steady hands" isn't stressed in surgical textbooks...it's assumed that this is a precondition for doing any kind of therapuetic work.

CT also doesn't disregard affective primacy. It says nothing at all about what comes first, it simply says that cognition is very influential and that cognitive restructuring can have a huge impact on emotion and behavior--regardless of where it resides in the "firing order".

And finally, your Pavlovian example of the tone/shock association with the dog isn't really analagous because dogs don't have cognition--at least, they don't have the kind that can be influenced with either direct disputation or Socratic dialogue. (And when you think about it, this isn't the kind of fear most people come into therapy with. I fear flying in airplanes...not because I've been in thousands of airplane crashes that reinforced this fear, but because I have an irrational fear that my plane will crash.)

I think your main point that not every kind of therapy works for every person is a very good one. I think that's too often ignored by therapists of all varieties.