Eugen,
Bell's theorem has been demonstrated to be an inaccurate characterization of the real-world, not merely via theoretical analysis, but by actually constructing 1,000,000 peculiar, entangled, classical entities, then measuring and analyzing them as per a standard Bell test, to reproduce the Bell correlations.
The demonstration exploits the fact that Bell's theorem assumes that the entangled particles being measured are identical. Consequently, Bell's theorem has nothing at all to say about any real world, in which "Mother Nature" does not know how to manufacture such idealistic entities.
Your "Modal assumptions" begins with the statement that: "With polarization measurements, photons can choose one of the two perpendicular polarizer exits."
My point is that, in the real world, "the two perpendicular polarizer exits" are not being chosen by the photons, at all. Instead, they are being chosen by a threshold decision process, that is "quantized" and consequently treats all photons as being in either an "up" state or a "down" state, even when any actual measurement, yields an "in-between" state that is neither "up" nor "down". That is what "quantization" is really all about - this was the most fundamental insight discovered by Shannon's Information Theory - measurement errors can, in effect, be entirely eliminated, via a properly constructed, "quantized' decision process.
In other words, it is the nature of the "quantized" decision process, rather than merely the nature of the entities being "measured" that is the cause for the peculiar effect. To put it bluntly, quantum theory is not describing any "state of matter" at all. It is merely describing that statistical behavior of a very specific type of "quantized" detection process, that is associated with the Born Rule.
Concerning MA4, your paper states: "Model assumption MA4: If the context of entangled photons, given by a polarizer setting, is changed to the setting of the other polarizer, then for this context we replace δ in all equations with -δ."
In this regard, you ought to reflect very carefully about the statement in my paper, that "Line number 82 in the script is commented out. If it is uncommented and executed, the "Quantum Correlations" disappear. Even the noise is now identical, except for a sign, for each entangled pair; causing even the bit errors (bad polarity decisions) to now be perfectly correlated."
The effect that line #82 has, is to replace one member of an entangled pair, with an exact, pixel-by-pixel, identical (negative) copy of the other member, instead of retaining the original, in which the entangled pairs are only "statistically" identical rather than being "exactly" identical. That simple substitution changes everything; in one case, Bell's assumption of exactly identical particles will be met, but not in the other case - and that simple change in the nature of the inputs ("identical twins" versus "fraternal twins) entirely changes the nature of the observed correlations between the entangled pairs.
Rob McEachern