Dear Tim,
In the interest of full disclosure, I should probably mention that when I sent you the 130 page precursor to my essay, detailing the local spin model and the Energy-Exchange theorem upon which my model is based, I did not realize that you had written a recent book explaining Bell's theorem and thus my model conflicts with your book. I know this would not cause you to obfuscate or distort my arguments in any way, but readers of the comments can at least realize what is at stake here.
You are apparently implying that Bell did not have any physics in mind, or any quantum mechanical eigenvalue equations, and made no physical assumptions in analyzing EPR, Stern-Gerlach, spin, and QM correlations, despite that he discusses all of these in detail in his papers, and despite Bertlmann's statements to the contrary. Readers can decide whether this makes sense.
Your first claim above was that my energy exchange model is global, but as you no longer mention this I assume you now realize it is a local model.
So let me simply use your terminology. You state that Bell merely assumes an experiment in which there are two outcomes, outcome 1 and outcome 2, plus a probability distribution for outcomes. No constraint should be imposed other than local causality. That is exactly what my local model does.
One experimenter, Alice, selects control variable a, and the theory ('any' theory, as you state) should yield either outcome 1 or outcome 2, based on actual EPR splitting observed. Any model actually based on physics will use her setting plus the local physics (denoted by lambda) to produce outcome 1, denoted by +A(a, lambda) that belongs to class 'up' or outcome 2, denoted by -A(a, lambda) that belongs to class 'down'. As is both obvious from the SG data and according to a standard QM text at the time, each class is "statistically distributed over a somewhat extended range". [See p. 3]
Bob's remotely operated experiment also yields two similar outcomes +B(b, lambda) and -B(b, lambda) which are based on the physics of the (any) local physical theory under consideration.
Then, as you note, Bell takes the correlation between the outcome of the two sides (in pairwise fashion) using the standard formula for expectation values. The standard formula, applied to my model's local outcomes, and plotted against the angle between Alice's and Bob's control settings, yields the top figure on page 7, exactly as predicted by QM. This is the cosine curve -a.b that Bell claims to be impossible for any local theory.
Now you ask where in his theorem do Bell's constraints appear. They appear in his first equation (1) where he states that +A(a, lambda) must equal +1, and -A(a, lambda) must be constrained to -1. There is no valid reason for these constraints, as they have the effect of throwing away the actual physics applied by (any) physical theory. I have explained why he erases this information in my essay and in more detail in my reference [4].
It is Bell's constraints, imposed on any local physical theory, that results in failure to match QM predictions. My local model does produce QM predictions.
Your use of 'toy model' is also incorrect, as toy models are 'reduced dimension' models. My model is full 3D and is a real physical model. Your statements about 'spin up' are also misleading, as I have explained in detail in reference [4]. In short, my model does predict quantized outcomes for spin, but the outcome of the experiment is not a direct measurement of spin, but a position measurement that reflects the physics of the spin scattered by the inhomogeneous field. Failure to recognize this has led to 50 years of non-intuitive nonsense about non-locality.
Regards,
Edwin Eugene Klingman