I am quite in tune with the idea of a spin at the speed of light. The electron could even be made up of standing wave photons. Being at the speed of light, just like the photon, the electron spin can neither accelerate nor decelerate.
The high rotational speeds of a disc magnet should increase/decrease its magnetic field, as it adds or subtracts turns/s to the spins.
But the difference cannot be significant: even with thousands of revolutions/s, the rotational speed of a magnet would be totally negligible compared to
the angular velocity of a spin considered as a current loop, of the order of 7.77×1022rad/sec (see
https://www.scirp.org/journal/paperinformation.aspx?paperid=97706).
In this paper, we also learn that
the magnetic field inside the spin would be of the order of 8.3 × 1013 T, which makes it completely illusory that an external field of the order of a few T or less has any significant influence on it.
I therefore share Broli's view that "any torque you apply on the electron spin will couple back to the bulk material", but that this coupling in the case of a magnet rotation is considerably below any expectation of experimental measurement, by more than tens of orders of magnitude.
The idea of playing on macroscopic elements and mechanical parameters to act individually on spin and modify its properties seems to me to be doomed to failure. I see spin as a perfectly stable entity that can only be an intermediate vector to act on something else.
Another example of the huge gap between the macroscopic and the microscopic. The acceleration of the electron is a source of electromagnetic radiation. Therefore, if a charged sphere is rotated, it will radiate. And as long as we remain in the qualitative, we can believe it. But when you go to the quantitative, everything collapses.
The acceleration of an electron in a field as weak as 1v/m is of the order of 20,000 billion g. Obviously it would reach the speed of light over an infinitely small distance. This is why relativistic corrections must be applied to its mass, which reduces its acceleration.
Can we seriously believe that with the centrifugal acceleration of a few g of the charges on a charged sphere, the electrons will radiate like those in an RF current?