[meroes]

The wavefunction is nonlocal. But if it's a complete descriptions of reality, it still provides no way to communicate FTL, and there really is baked-in randomness. This is because local measurement outcomes on half an entangled pair, which necessarily use macroscopic, local apparatuses, can never be predicted to arbitrary accuracy in advance (unless entangled partner is local too). You will see random outcomes locally, even though the distant entangled partner's outcomes will be perfectly correlated upon classically going and looking at the distant outcomes. If the WF is not complete, then there is a deterministic, FTL underlying theory, which may or may not allow FTL signaling in spacetime. Those are the only ways to satisfy Bell's results with his modest assumptions. Those assumptions are sidestepped in things like MWI and superdeterminism.

(Assuming the same assumptions as the Bell tests)

(And assuming things like "measurement" and "macroscopic" have well defined meanings, which Bell had problems with too)