As always, an impressive essay, full of relevant examples, and focused on the overview. You begin by noting that our organic computational systems have abilities different than computers and suggest that one proper use of these abilities is to
"keep looking for the boxes outside the boxes which the boxes came in."
You also note "most believe no classical rationale [re Bell] is possible", and discuss the herd behavior that shuns any deviation from conventional wisdom so that few dare to break ranks.
Your '3-filter' example beautifully summarizes the nature of propagation, and is followed by Zeilinger's very important observation that:
"...Light always has the polarization state given by the last polarizer and has no memory of its earlier history."
That is essentially the insight that I employ on Stern-Gerlach, i.e., the dipole moment aligns with the field, versus the QM view that it is only "the component in the field direction" that is being measured. But the 'unaligned' initial spin possesses precession energy, and there is no precession energy when the moment is aligned with the field. Where does that energy go? Obviously it goes into the kinetic energy of the deflection from its initial momentum. Thus the energy of precession when the initial spin enters the field is exchanged from the precession mode (rotational) to the deflection mode (linear) and this determines the scattering of the particle in the inhomogeneous field as indicated by the position measurement.
Bell assumes instead that the particle simply continues to precess all the way through the apparatus and thus has no way to produce any but a constant deflection, normalized to ±1, and, as this is compatible with the basic Quantum Credo that "measurements yield eigenvalues", he imposes unwarranted constraints on the problem. Much of my essay is concerned with why he does this.
In a nutshell, he is tricked by math, assuming that, because the Dirac fundamental eigenvalue equation "looks like" the Pauli provisional eigenvalue equation, which is valid only provided a constant field is present, which it is not! You note on your page 2 Wittgenstein's analysis that "looks like" was the reason why all 'knew' the sun orbited Earth. Today it is the reason why all 'know' that no local model can produce the quantum correlations, which my local model does produce.
The whole argument is subtle, or it obviously would not have evaded physicists for 50 years, but now it is a fundamental part of the QM faith, and it is dangerous for the faithful even to discuss it, judging from comments on my thread [or lack thereof].
Anyway I know you have been concerned with this problem for years and I hope you will find the time to study it until you understand it, or will ask me questions otherwise. Last year I promised you a new approach, but it took longer than I thought to work out all the details.
My very best regards,
Edwin Eugene Klingman