Mauro,
I've reposted this reply to you from Bill McHarris's blog.
No 'changing meanings'. Theorems are indeed theorems, but they're all included in the greater 'theorem' that all science is provisional and no 'absolute' proof of anything exists. Bell uses assumptions just as all theorems do. Even the most solid foundational 'Laws' of Physics can be violated. Look what happens to Snell's Law at kinetic reverse refraction - the nonlinear 'Fraunhofer refraction' appears instead!
The measurements are detector angles and 'positions' along the x axis of a cosine curve distribution between 0 an 180 degrees. Consider my torii as entangled particles translating along the polar axis with opposite spins. They meet detectors as 'planes' A and B tilted at varying angles (or tilting donuts if you prefer!). 'Detection' is of the interaction point at A and B, which is say in the top half ('up') or bottom half ('down').
We now have another 'dimension' that Bell did not assume existed. We can easily show that when A and B are parallel the results are opposite, and when anti parallel the results are identical. But half way between, when A or B are vertical the donuts hit face on so the result up/down is at maximum uncertainty! But over many samples it is of course ~50%.
Now the killer; When intersecting at 90 degrees, tilting the detector say 5 degrees will have virtually no effect on the 'position', but when face to face, a 5 degree tilt angle has a major positional effect! So 30 and 60 degrees give results of 75% and 25%. This is Malus' Law in action, and reproduces the predictions of SR at EACH detector (just as von Neumann proposed) as well as when correlated between the two.
All this is as published in my essay and expanded in the Blog. Aspect did find this "orbital asymmetry", but with no theory to fit it to he discarded that particular ~99.9% of his data! (only discussed in his follow up French paper).
This is very consistent with Prof McHarris's findings and I believe Gordon Watson's essay, with similarities with Ed Klingman's. I'll re-post this on your blog so you don't loose it. Do ask any questions or give views on mine.
Very best wishes
Peter