In particular, standard quantum mechanics gets results of experiments by collapsing the wave function. That is, it takes something which is universe spanning and it collapses it to something, with the collapse based on what a physicist does in a laboratory. That, of course, should seem rather absurd. To get around that, you have to add something either particle positions, a matter density, flashes, etc. You need results to experiments in a theory. Without something (and collapse from experiments is adding the collapse mechanism + external experiments, whatever that means), you do not get results. So the theory is a complete failure.
Only many worlds is not non-local (it adds a matter density by integrating the wave function appropriately), but they escape it by not having results of experiments at the time that history records there were results. I think. That theory makes me a little dizzy at times.
Also, the hidden state of BM is not hidden. We have a basic sense that stuff has positions. That may be an illusion, but Bohmian mechanics says that we can choose a theory that does not have that be an illusion. The thing that is hidden, that is deduced from the behavior of stuff which we see, is the wave function.
What BM explains is why it seems that our world is particle-like and why we have singular outcomes to experiments. All done naturally. Sure, the nonlocal nature is a problem that we are still trying to understand, but so what? No one has got the full story yet and so BM is viable and gives a great intuition about QM (the standard QM (operators, collapse) is a derivable formalism, like thermodynamics from Newtonian Mechanics).
I'm all for trying to formulate QM in a way that doesn't offend basic sensibility, but I'm not convinced this is it. I don't think deterministic-but-nonlocal is better than nondeterministic-but-local.