Alright, but what I really wanna know at this point is what Brian Greene is referring to when he says there can be some time-delay between the collapse of the wave function of A and the one for B. I thought a lot about this yesterday, and the conclusion I got (which also solves my problem) is that the wave function collapse doesn't actually happen at a particular location in space.

I mean, the wave function collapse is just a name of a reduction of the spin wave function from being in the superposition UD+DU to being of the form UD or DU. But this wave function is a ket vector in Hilbert space, so it simply doesn't exist in ordinary 3D space.

What Brian seems to have done is identify the location of particle A, say, and the time of the wave function collapse as the EVENT where the collapse happened, and likewise for B, and in that case, then sure, there might be a time-delay according to the rules of special relativity.

The important point here is that all the rules of special relativity only apply to EVENTS in spacetime, which is a collection of a location in space and a moment in time. But if the collapse has no location, then it doesn't happen in a particular event either.

Therefore, it can simultaneously collapse for both particles for all reference frames, without violating special relativity. This would also solve the entire problem. But if this is all true, then why did Brian Greene, a leading theoretical physicist, say that there must be a time-delay?