Richard,
"Hidden" assumption meant hidden from Bell too. He agreed; "there are many other ways of creating the troublesome correlations." But 'avoiding assumption' means precluding no possibility. I've had Bertlmann's socks for years (he didn't complain - I think he had an identical pair) and I know Bell tried hard to avoid assumptions. However, just making "no reference" to particles or fields doesn't necessarily accomplish that aim. He missed a possibility so used a hidden assumption. The proof is simple. He only accounted for ONE of these options, not BOTH;
1) That the 'entities' arriving at A and B are different and paired, so cannot be affected or changed by the 'preparation', detection or 'measurement' interactions.
2) That the entities are similar but with two (and opposite) orientations, the measurement 'results' of which can change with state of interacting particle/field.
Bell 'missed' the 2nd possibility because QM's 'singlet state' description also seemed to preclude it. His own proof then didn't apply to it.
Now note; for 1) a) No such change is detectable by weak measurement. b) The anomalous 'rotations' data from timed pair experiments is inconsistent with option 1.
However, Option 2) is fully consistent with both timed pair experiments and weak measurement data. It also employs verified optical effects (non mirror symmetry of AOM and electron spin-flip. By any objective analysis 2) is then correct. The reason it's only detectable by timed pair analysis is due the nature of randomness; even if ALL spin findings are reversed the 'north/south' pole distribution is identical at 50:50.
Statistical analysis is still helpful but now applied in a different way, so your apparent tendency to partisan voodoo on that basis is unnecessary. I know this may be a shock to a brain not rigorously trained in the scientific method and used to trusting familiar assumptions (i.e. most!) but such 'belief' has zero value in determining truth. Just thinking clearly about 2) will reveal it's veracity.
Best wishes
Peter