Dover
Adding mass to the spindle helps to stabilize it.
To repeat, for a given resonant system, extra mass lowers the resonant frequency and reduces the amplitude of this resonance. This is F=ma again. Force the same, Mass up, means acceleration must go down.
I agree that adding too much mass is potentially a problem. Where we disagree is at what point this extra mass becomes a problem. Pivoted arms, as seen by the cantilever, have high effective mass due to the head shell offset. Yet nobody seems to worry about this. I would also suggest that anti skate is significantly more dileterious to the cartridges health.
Placement of magnets..... Below the bearings natural resonant frequency, it is effectively loose, above it is nearly rigid.
So the thought experiment is where is it best to place a restriction to movement which creates a pivot point in order to minimize unwanted movement of the wand. If the mag is placed wand end, the bearing clearance allows wand movement at frequencies below natural resonance but not as much as when the mag is placed at the counterweight end. This is not a bearing stability issue, it is just optimizing the bearing geometery. Like I said this the same principle employed with inverted bearings in TTs.