Basically people who practice CBT think that people who practice psychodynamic/existential-humanistic spend too much time analyzing things that may not matter with less direct evidence that it works. The other half(myself included) think that CBT is too much of a quick fix, and that yeah, you may want to have 6 sessions of skill building to solve your problems, but that will only provide a short term solution.
The reality is probably somewhere in the middle, and the skills taught by CBT are probably important. And if you have poor coping mechanisms or antagonistic thought patterns CBT is important, but it won't touch the underlying issues.
The funny thing about this article and the talk about the "placebo affect" is that to a certain extent many types of therapy are about integrating experiences: coming to an understanding about events in your life and the context in which they take place. The other thing that is linked most highly to success in therapy is the strength of the relationship between therapist and client(with a few caveats. The research shows this relationship should embody a few things and obviously it could be strog and unhealthy if there's something unethical going on).
But basically, talking about the placebo affect in mental health is kind of a funny thing. There are studies that look at placebo affects to certain therapies, but in general, since the goal is to feel better, if you think it worked and your supposed to feel better, should I come back and say don't feel better?(barring concerns about it being a short term effect). When the goal is to think you feel better, how(and why) do you suss out the difference. What I've always wondered, is do we think about the placebo affect the wrong way? Instead of seeing as something that disapproves efficacy, should we think about it as a power we should try and harness?