At the moment, there is no canonically useful model of the mind that is established in the consensus of those who study it. You are writing this as though psychoanalysis has been replaced by some better theory that explains more of the phenomena that it addresses, but no such theory exists. One of its main contenders, evolutionary psychology, does not actually provide a model for the mind, just the conditions under which it developed. Other models have offered piecemeal models of isolated effects such as cognitive dissonance, many of which are subsumed by psychoanalytic models anyways. The point of Kandel referring to it as the "most satisfying model of the mind" (at least in 1999) is because it addresses a territory of phenomena that still does not have a clear alternative. There is no analogous "geocentric system" against which psychoanalysis looks Ptolemaic. From the above reference to work, clearly psychoanalytic models are not useless, and hence researchers continue to find reasons to stick to it.
I think you are conflating Freud's model of the mind with a couple of particulars. The id/ego/superego is not particularly central to Freud's work until it makes an appearance as the "second topography" in the 1920's. The most important part of his model involving unconscious processes are almost taken for granted by psychology today after trying to soft shoe around the idea for half a century via reductive behaviorist models that fail to substantially account for anything complex like language, culture, thought and memory.
Even if we address the second topography, why is it that it "probably doesn't exist?". Is it because Feynman doubted it? It would certainly be dogmatic to suppose that this suffices to justify this claim. Solms et al. have done a lot of work showing correlates between the model of the ego and the id and functional neuroscience, so there is work to show that they very well may exist. You cannot just assert that these things don't exist because you or Feynman don't personally buy into them. None of these researchers are taking Freud as gospel, as you seem to be taking the words of Feynman. Denying that psychoanalysis is useful means engaging in those actually working on its ideas, or even engaging in Freud's ideas substantially, and making clear arguments for why this should be treated as useless.
I think you are conflating Freud's model of the mind with a couple of particulars. The id/ego/superego is not particularly central to Freud's work until it makes an appearance as the "second topography" in the 1920's. The most important part of his model involving unconscious processes are almost taken for granted by psychology today after trying to soft shoe around the idea for half a century via reductive behaviorist models that fail to substantially account for anything complex like language, culture, thought and memory.
Even if we address the second topography, why is it that it "probably doesn't exist?". Is it because Feynman doubted it? It would certainly be dogmatic to suppose that this suffices to justify this claim. Solms et al. have done a lot of work showing correlates between the model of the ego and the id and functional neuroscience, so there is work to show that they very well may exist. You cannot just assert that these things don't exist because you or Feynman don't personally buy into them. None of these researchers are taking Freud as gospel, as you seem to be taking the words of Feynman. Denying that psychoanalysis is useful means engaging in those actually working on its ideas, or even engaging in Freud's ideas substantially, and making clear arguments for why this should be treated as useless.