VPI Classic tweak

How else could the speed be different? The pulley is machined to a different diameter at each of the pulley positions.

Ok, so that I understand you, you maintain that each of the three grooves on the 33 rpm (and 45 rpm diameter) that are millimeters apart have a different diameter? As I stated, the pulley is manufactured to such high specifications that the miniscule differences, if any, should have an absolutely negligible effect on the platter speed. I believe the differences in speed are caused by different belt positions relative to the platter bearing; i.e., the pull is different when the belt is higher or lower on the platter.

I believe the position on the platter is irrelevant, the platter is machined to the same diameter up and down its sides. I believe the pulley is machined intentionally with a larger diameter on each groove progressively to accommodate those enthusiasts without a SDS. This is the only way they have to compensate for AC frequency differences, which is how speed is determined on an AC motor.

It totally makes sense, if true, and would perfectly explain the speed difference among different pulley positions. However, I'm still unclear why using two belts would not result in a net speed that could be adjusted by the SDS since both belts would be subject to the same forces with each rotation.

The belts will be binding, one belt will try and turn the platter faster in each rotation than the other. Therefore one will have to slip. The belt with the most torque will probably win. Hard for me to visualize how this could be an improvement in sound.