Halcro,
I think you're on the wrong track ...
Simple physics dictates that speed must "swing" above and below 33 1/3 rpm in order to average out at 33 1/3. A speed change to below 33 1/3 HAS to occur first, before it can be detected and then corrected by applying more torque. If a drive mechanism does not "make up for lost time", the timeline laser would slowly, but constantly drift. The fact that these speed changes are very very small make it difficult to understand, but they have to be there in order to be corrected. Without any servo circuit measuring and correcting speed changes, the speed-torque curve of the electric motor would have to be a vertical line ... and there simply is no such thing.
You are right, it is very problematic to ceorrect speed in a way that our hearing tolerates it. The very high torque drive system of a Technics SP-10 MK3 or a JVC TT-101 sound much different from a low torque direct drive, but they both are able to maintain correct AVERAGE speed of 33 1/3 and the timeline laser will not show a differnce.