To the extent that thought experiments suggest real experiments, they are indeed valuable. I disagree that they can 'reveal ... results', however. For instance, Einstein's thought experiment was designed to 'prove' that QM was wrong. Obviously, that result was incorrect.