Lewm: "IMO, the rim drive is the worst of both worlds, not the best of both. Mechanical vibrational energy from the motor is transmitted right into the platter with no belt to isolate one from the other. At the same time, the typical rubbery contact point between the drive wheel and the platter is constantly trying to rotate the motor in the opposite direction (per Newton's 3rd Law of Motion), and flaws in the O-ring result in mechanical noise and speed issues."
I agree that it's the worst of both worlds. My main issue with it is that it takes away idler drive's advantage keeping the metal (motor pulley) to metal (platter) relationship with rubber (idler wheel) interface in between. In an idler drive the rubber wheel's size does not affect the speed (at least not much) because it's only an interface in preserving the relationship of two true round rigid metal circles of the pulley and platter. Idler wheel is compliant and will have pressed spots touching metal and yet the speed does not change. But a rim drive must have true round rubber wheel to have speed accuracy because that is also the pulley. Just imagine a non-true round pulley in belt-drive and idler drive. Disaster. Rim drive's rubber too hard will not grip the platter and too soft will deform the round shape and affect speed. One might say rim drive has the advantage of transferring the effective torque (including motor noise) more directly to the platter, and it can be the reason people like the sound. It demands a very quiet motor too. For me, I rather stick with idler drive between the two systems.
Just for the record, I believe all three systems (belt, idler, direct drives) can sound good. It's the execution that matters.