I haven't had difficulty grasping your point that it is not useful to conduct research which tries to prove that an Ericksonian approach to therapy is better than, for example, a cognitive-behavioral approach. In fact, I agree. I just think that this is a straw man argument because it implies that the only kind of psychotherapy research available consists of efficacy studies.
To address your question about whether follow-up data would tell us anything about the case. It is true that data provided by long-term follow-up on one case would be limited to the patient's perspective about what worked or didn't work. However, routinely gathering long-term follow-up data on all cases allows clients to give feedback about their perspective on therapy and allows clinicians to produce data that are understandable for non-clinicians. For example, of the clients treated for migraine headaches, __% noted a reduction in symptoms. I would argue that these statements have real world importance to clients and other people who pay for therapy. Of course, the data have many, many limitations but that does not mean that such research cannot contribute to a clinical understanding of patient outcomes.
While I am not refering to your statements in particular, I think it is most unfortunate that the field of psychology has developed such a split between researchers and clinicans when we should be working together to show how therapy can benefit those not traditionally served. When communicating with students clinicians can be subtly or not-so-subtly disparaging about what research can offer and this in turn reinforces an artificial dichotomy between science and practice.
Thanks for allowing me to make this point. It is a worthwhile discussion but perhaps there are others on the forum who want to contribute and I will leave further to debate to them!