According to the paper, Bohmian mechanics violates assumption (Q) (the first one), which is not exactly "consistent with QM", but rather "If an agent knows the state of a quantum system S and knows that a measurement of x applied to the state yields ξ with probability 1, then the agent knows that x = ξ".
Bohmian mechanics violates it by not applying to arbitrary subsystems, but only to the universe as a whole (i.e. parts of the universe cannot be treated as quantum-mechanical systems themselves, because they don't have their own pilot waves)
I find that to be logical, as it seems like it would be incredibly difficult to dis-entangle an observer from the minute fluctuation and relationships with every other bit of energy in the universe.
“The etymology of the word ‘universe’ can be traced back to the use of the Old French univers, in the twelfth century, which derives from the earlier Latin universum. This word is created from unus, meaning ‘one’, and ‘versus’, the past participle of the verb vertere, meaning ‘to turn, rotate, roll or change’. So we have a literal meaning of everything ‘turned into one’ or ‘rolled into one’.”
from John D. Barrow’s “The Book of Universes.”
Sounds like Bohm is taking the 'uni' in universe seriously, while the paper starts by assuming there is no fundamental unity in the universe.
Parent is correct, Bohmian mechanics violates (1), but we already knew this. Bohmian mehanics allows for systems in quantum non-equilibrium, which is a unique concept to Bohmian mehanics. Unfortunately we don't know how to create such a state to test for it.
Bohmian mechanics violates it by not applying to arbitrary subsystems, but only to the universe as a whole (i.e. parts of the universe cannot be treated as quantum-mechanical systems themselves, because they don't have their own pilot waves)