I'd like to add something regarding local causality. In a sense, this is a discussion about continuity. At the risk of oversimplifying, we could say that QM is a theory that, given a past event, allows us to calculate the probability of another event occuring in the future. These two correlated events are obviously "real", we know they occur. For this calculation, we introduce the quantum state that unitarily evolves from the initial time to the final time. To relate one thing to another (the actual outcomes and the unitarily evolved quantum state), we use the Born's rule.
So, the questions are: (i) how to interpret what happens between these two events, i.e. the unitary evolution of the quantum state, and (ii) how to smoothly connect this unitary evolution with the appearance of the detector's clicks. Many people don't agree with the idea that nothing "real" happens between the two events, so they try to provide the theory with some ontology that "fills" that gap. In many-worlds, the quantum state is considered as the only thing that exists. In Bohmian mechanics, along with the unitary evolution of the quantum state, there are the particles that (continously) move from the first event to the second. In GRW-like spontaneous collapse, a sudden state reduction explains the transition from spacially distributed wave functions to well-localized detection events. On the other hand, some people try to interpret the theory in its minimal form, without adding anything to it. This is how some epistemic interpretations work. We could mention Copenhagen-like interpretations or information-based ones, such as relational quantum mechanics. In these kind of interpretations, as
@Morbert said in
post #81, reality is associated more to the interaction between systems than to systems themselves.
In my view, the above distinction allows to understand how each interpretation addresses nonlocality. For interpretations that attempt to fill the gap between two correlated events with something ontic, a non local mechanism is usually considered as an explanation for non local correlations, as in Bohmian mechanics. For the interpretations that consider interactions to be the only real things, there is no fundamental causation that propagates the cause in the past to the effect in the future in a smooth and continuous manner through spacetime. That is, the discreteness of the QM realm is accepted as fundamental.
Furthermore, for the second kind of interpretation, even when the mathematical tools employed to calculate probabilities for events' correlations are not considered "real", relativistic causality plays a fundamental role in this calculation, by evolving the state of the system between the spacetime points of the two "real" events. In other words, relativity is correct for calculate these transition probabilities, but it has nothing to do with possible outcomes becoming real outcomes.
Lucas.