I'm not at all surprised about Chris' experience with the granite slab, but I see it in slightly different terms than he does. I don't think that removing the compliant feet gave the slab a 'pathway' for resonant energy to 'dissipate', so much as the full-surface contact between it and the underlying wooden rack-shelf provided very much more damping, needed to kill its strong ring. Think of it this way: Wouldn't the same thing have basically happened if he had placed the shelf *on top* of the slab, only to a lesser degree due to the shelf's lesser weight?
My non-suspended TT is placed on a Symposium shelf atop compliant footers, but that shelf is not massive and is fairly well internally-damped and so doesn't ring much on its own, unlike the granite slab. High mass alone is not always our friend in this area, especially depending on how its shaped (think of an equivalent mass shaped as a block and as a slab, for instance, and then imagine a gong and an anvil made of an equal mass of the same metal - the gong rings much lower and longer, the anvil much higher and quicker).
All vibrational pathways are two-way, and I don't want my unsuspended 'table not to be isolated from floor vibrations or from airborne vibrations that are transmitted to my audio rack. If everything is rigidly coupled, then acoustic feedback becomes a real possibility with an unsuspended 'table at high playback volumes, especially in the case of a suspended wooden floor.
Chris mentions 'dissipating' self-generated vibrational energy from the TT on one hand - and how compliant isolation can supposedly thwart that goal - but then goes on to describe how Teres intends to damp this vibration internally through their chassis design. My belief is that these two statements are somewhat contradictory, and that for dealing with energy created at the bearings and stylus/groove interface, damping is the only way you can realistically go (through the use of nonresonant TT construction - which since TT's are rigid means damped, either by using relatively self-damped materials like acrylic or by constrained-layer mixing of unlike materials - and an appropriately compliant full-contact mat with clamp, though some will argue that).
To me, the reason non-suspended TT's can work well in the real world, and the reason less-ambitious suspended designs might not always be unequivocally superior, has largely to do with damping. Many less-expensive sprung-chassis designs are essentially undamped or only rudimentarily damped in their suspensions, so that even though they are resonantly 'tuned' as low-pass filters, this is like a choice between a car with no springs, and a car with springs but no shock absorbers. Clearly, though it will be more complex, cost more, and require more precise engineering, the best ride will ultimately be gotten with both springs and shocks, and I do agree with AJ that a premium TT ought to be both fully suspended and that the susupension needs to be effectively damped.
But on the budget end, where I live, I agree with Chris that unsuspended might often have the potential to be best in practical terms, although I'll demur about rigid coupling as a rule. I think an unsuspended 'table that's internally well-damped, coupled with a non-resonant rigid shelf that's mounted on appropriately-chosen compliant footers, can emulate to a useful degree both the low-pass isolation and overall system freedom from resonance that you would expect from a more expensive suspended design. In fact, if we disregard for a moment the question of where you put the motor, it seems to me that there might be little intrinsic difference between a 'table that carries its own sophisticated damped suspension, and setting an unsuspended 'table on a platform such as a Vibraplane that can perform largely the same function. But - perhaps unlike the manufacturers under discussion her - I've never put that proposition to a practical test, and neither have I ever owned a competently suspended/damped premium TT.
