Use of a capstan, etc, or any strategy that increases the area of contact between the belt and the platter is a good way to combat "belt creep", the bete noir of belt-drive tt's. This is why I am skeptical about 2- and 3-motor arrangements; of necessity, the platter to belt contact area is reduced in both cases, even compared to conventional single-motor arrangements. This is also why I really like the Artemis SA1 turntable, in principle. (Never heard it.) It uses a single motor and a capstan so that the belt (looks like tape) is nearly completely in contact with the circumference of the platter.
Dare I say that a device like the KAB might be superior to the Timeline for detection of very transient deviations in speed? OK. I said it.