Thanks Peter,
Critique follows re this:
"If magnet A is rotated; the particle deflects down. The particle at B then MUST go up. Therein is the problem. If magnet A is reversed at the last moment, how can B know without "spooky action at a distance"?
Or this:
"At identical angles we find opposite deflection (or 'particle spin'). At 'opposite' angles, we find identical deflection (spin) direction. But that gives the big issue; Reversing magnet A at the last moment DICTATES particle B's deflection/ spin state!"
Critique: IF magnet A is reversed at the last moment (or at any time) during this run so that it is antiparallel to magnet B, what makes you think that B has to know anything? It does not.
Experiment (a): If magnet A were left unchanged (ie, being parallel to magnet B), let the paired result be, SAY: [Alice = Up; Bob = Down].
Experiment (b): If (instead) magnet A had been rotated PI radians, to now be antiparallel to magnet B, the comparable paired result would be: [Alice = Down; Bob = Down].
That is: Alice's rotation of the magnet by PI radians reverses Alice's result. BUT it has NO influence whatsoever on Bob's result.
Reason: Let Z denote the EPRB experiment and let p(.|Z) denote a probability conditioned on Z. Then, surely by your own analysis:
Experiment (a): P(Alice = Up| Z, Bob = Down) = cos2(0/2) = 1.
Experiment (b): P(Alice = Down| Z, Bob = Down) = sin2(PI/2) = 1.
With best regards; Gordon