we can say the same for Flat Earth theory: it works great in everyday life, but it was a scientific mistake.

I think you are misusing the words. We don't call theories a "mistake". Theories can be "wrong", if their predictions contradict experimental results within the area they claim to be applicable. It doesn't make them a "mistake". A "mistake" is an action that leads you away from your goals. For instance, you can say "Building a theory of everything based on Lorentz invariance is a mistake". Building a theory of everything here is the goal, and building it based on Lorentz invariance is the action, which, according to the author, does not serve the goal well. Calling a scientific theory or parts of the theory a mistake does not make an intelligible sentence and just sounds crackpottish.

If you imply somehow that the assumptions of Lorentz invariance or postulates of quantum mechanics are wrong, it is important to specify the context, that you really mean that they are wrong in the context of the theory of everything. I am not sure that even putting it within the context would make it true, though. It may turn out that the principles of QM or GR are correct (see string theory). There is simply no evidence at this point.