Try a different turntable?

While I agree with Schipo on retaining the VPI (and its underscoring quality), you do seem to be in a position to try other tables just for the sake of it so why not. Maybe you can convince a local dealer or two to let you trial some tables in the other system, and with the wide open resale market, you could probably buy and sell some of the fad alternatives (ie direct drives and rimdrives) being touted. Who knows you may become a debunker in the process...
However, If you want to just try something similar in design and concept, maybe a used similar vintage SOTA Star, however they still have a largish footprint and real heavy.....

Oh did I say debunker? Just think about how an idler wheel functions or a DD for that matter. Virtually all of these designs are heavily modified in plinth, etc, etc, etc. Also note the size of these nantais designs and the Op's space requirements.

Sure, in both cases the motor is more or less solidly coupled directly to the platter either through the DD shaft or through the quite solid coupling wheel. In belt drive it is decoupled by the belt., why would a properly designed belt drive have speed stability issues??? where does this myth arise from?

Are you telling me the drag force of a delicate stylus at say 2 grams would alter the speed of a moving disc at 4000 gms or 10,000 grams (roughly the weight of my platter). I would bet the friction required to do that would tear the tip right off the stylus.
Do you really believe this?

I'll stick my neck way out and suggest you were-preconditioned to expect that and you wanted to hear it (without knowing which tables you are talking about), and I am not trying to be insulting, just practical. I do think it is a credit to some of the upgrade/rebuild guys working on Lenco's and Garrards that they can apparently remove a lot of the foibles (wow, rumble, flutter), although I have yet to stumble on any reviews where 'actual measurements' were made using conventional testing equipment.

quote lewm: "The stator or magnet part of the motor is actually part of the inert platter itself or is firmly mated to it."
I assume this is when it is both inert and in motion, either way how could it be any more directly-coupled than that? The principal problem with direct drive turntables when they were introduced was 'cogging' created by movement between the magnetic poles of the motor. So they reduced the problem by adding additional poles to the motor. I think around 24 was thought to be enough, however that's still 24 cogs in around 2 seconds. I can think of 1 high-end DD turntable manufacturer only(Goldmund) and one other mass-producer that has a quite romantic following (more initially with DJs) and I understand they have announced they are going to discontinue building their tables (notwithstanding the vinyl resurgence). Their romantic resale value should rise even more accordingly.
As far as rim-drives go, this is mostly a market of diy/garage builders based around 40 year old designs
rebuilding salvageable $50 (c 1960)tables. Oh, I do hear VPI offers a rim drive option on one of its pricier setups. Either way, its a wheel, directly driven by a motor and the wheel makes direct contact with the platter.

Anyway, I actually am thinking seriously about trying a KAB 1200 myself in one of my systems.

Lewm, I think you may be mistaking me for someone else with your first comment directed at me, however I do still appreciate your job offer.
Although I do believe DDs came a long way with control of speed, it is not all that diffult a concept to understand why a belt drive in general is less immune to 'cogging' than a direct drive. Left uncontrolled, a direct drive would pass from pole-to-pole in a 24 pole motor approximately once every two seconds. with a roughly 38 inch perimeter circumference that would equal about every inch as well, and should be quite obvious to someone whose powers of observation allow them to visibly detect stylus drag on a 22kg platter. The belt drive motor spins at up to 600 rpm in some cases with a circumference of the spinning shaft of less than a cm in most cases. The perception of cogging is now in milliseconds and then it is insulated by the belt. That would be the main difference.

TTWeights is a CNC machine shop in Ontario, who about 2-3 years ago started making weight clamps for turntables, perhaps as a bolster in the downturn of the auto industry. Im guessing the owner is an audio enthusiast who picked up on the current trends, who knows if he has the knowledge to design a really good turntable. He has gone very rapidly from machining record weights to purporting to be a designer of high end turntables.

'Stylus drag' seems to be the principal theoretical benefit of rim drive designs. Personally I think it's hooey. If it was constant it would not be perceivable and to suggest a minute change in a track would alter the speed of a turning 22 kg massive platter is frankly not believable.

I'm certain I will enjoy the KAB if I do pick one up for a try. I might even take up scratchin'.

Cheers, Dude.