I have already said that there is nothing to "prove" (i.e. no proof) but that it is a necessary corollary of applying the heuristic of Occam's Razor(https://en.wikipedia.org/wiki/Occam%27s_razor).
> A more complex hypothesis isn't useful until you come up with a method of testing that will distinguish it from the simpler hypothesis.
No, the condition is even stronger; a more complex hypothesis should not even be considered until the simpler hypothesis fails for some data/evidence.