Upgrade from TW Acustic Raven AC-3 to what?

Laws of physics say that for every action there is a reaction, therefore any "give" in a bearing means you are losing the leading edge ( I think you will find the downwood pressure/energy of the minute stylus tip is surprisingly high ). For an air bearing or magnetic bearing to be completely rigid they would have to have infinite pressure - impossible. The only benefit of air bearings is that they are much cheaper than a decent mechanical bearing that does not induce vibration and noise, they dont wear out and they are more forgiving of poor set up ( soft bearing = soft sound ). Remember Enid Lumley of TAS - drop the air pressure until it fouls and then raise it slightly - this gives you the softest mushiest sound possible.
Lewm
I can only draw on experience - listening to the L07D demonstrated the later Goldmund direct drives were cogging, similarly when comparing the SP10 Mkiii against the Mkii you can clearly here the instability of the Mkii. The mkiii is the best dd I've heard - I'm just not totally convinced. I would not be averse to picking up an SP10 mkiii for experimentation to find out more.

Writing off air-bearing featuring TT's categorically because of an individual point of view how bearings in TT's work and interact with tonearm, plinth and underground is not at all an approach worth further consideration. While the SP-10 MK3 is a nice and well made DD-TT, it is nowhere near an estimated nor possible optimum in TT design. It is a 30+ year old design made for broadcast applications and following a certain, - then en vogue - principle of the day. As the idler-drive TT were 2 decades earlier.
An air bearing (working...) being less expensive than an average "true mechanical grounded" bearing ? What do you think the bearing in the SP10 MK3 did cost ? Whatever you guess now - its not half of that figure.
It is somehow funny, how - especially in TT design - everything is worshipped which tries to get around physics and real investemnt in material and financial resources. The one and only real clever approach in making a good cheap TT I have ever seen was Bill Firebaugh's initial design - it only suffered seriously from choosing the wrong velocity in its damping fluids.....

True mechanical grounding ? Tell that anyone in any laboratory working with microscopes - you're in for a good laugh and instant empirical proof that it won't work.

We will never see a lightweight TT nor an unsuspended one, with a platter less than 30-50 lbs coming anywhere near the point of closing the book on TT nor approaching its true frontiers.
As we all will see - as they will come and go in half years turn.
The same they have done so for the past 40+ years in high-end.

I can certainly understand the microscope analogy. However, microscopes don't create vibrations nor do they interpret them, so to some extent the analogy is not complete as a comparison. I know many of the isolation techniques used in audio look at those used for elaborate electron microscopes. Turntables are selective vibration detectors, however, with the vibration starting at the device. They also have to decide how much vibration to "keep" and how much to "filter". That is somewhat different from maintaining a clear visual field in a microscope.

Managing the vibration introduced by the tracking process is a different matter. This is not disputed in any way by me. However - the isolation from OUTSIDE vibration traveling into the TT - is essential, a true conditio sine qua non. The inherent vibrations have to be eliminated inside the system (here: TT) - the external vibrations have to be kept away from the system (again: TT...).
An ideal solution/scenery: different room to eliminate sound pressure effects PLUS isolation from building resonance by means of a vibration isolating platform tuned for below 2Hz resonance frequency - 0.5 Hz if possible.
Regarding the isolation from outside vibration TT's (at least the ones worth mentioning and claiming themselves to strive for the state of the art) do indeed have the very same needs as an electron microscope.
Avoiding having the very tiniest details smeared by low frequency outside vibration moving into the TT system.
Anyone not accustomed to this, I strongly recommend giving their beloved TT a listen after setting in on a Vibraplane or Minus-K with enough load close to the maximum allowance.
The following listening will answer all questions in a matter of 20 seconds.
I know it - I have watched the faces of various audiophiles often enough in those first seconds.
After that all critics turned to true believers in the simple laws of mechanic.
But - its kind of costly and won't work with some of the fancier "high-end" rack due to lack of guts (here: stability).
It doesn't look to good either.
A well loaded Vibraplane moves into the game very much like a solid full linebacker with 240 lbs+.
But after all - this is a game for real men.
And here once again the boys are separated from the men - as the later do what has to be done and not just talk.
Enjoy the journey,
D.

Isolating the turntable system from external and airborne vibrations is certainly essential. That creates one point of vibration attenuation and frequency reassignment from external influences.
A lot of the above discussions seem to bear on the issue of the various interfaces of the vibration detection system, itself and how they "drain" into this interface between the turntable system and the outside "world".
For instance, a theoretically perfect rotating platter with a perfect gravitational field on a perfectly frictionless air or magnetic bearing creates another point of isolation within the turntable system itself. Also, the more frictionless the bearing, the cartridge modulations and imperfections of the vinyl will create centripetal and centrifugal de-stabilization effects over time which can only be reflected back into the vinyl/cartridge system from the platter to the cartridge. I imagine this is why the Caliburn system unloads the mass but preserves a mechanical interface, to stabilize this interface and create a point of vibration drainage apart from the cartridge interface, itself.
Also, a linear air bearing tone arm, also theoretically frictionless, creates another point of isolation from the internal interfaces of the turntable system. Although this may minimize tracing distortion, it creates two different effective masses and another isolated vibration interface. Unless the damping from this system is perfect, the cartridge sees reflected vibrations from the tone arm system itself without an external source of vibration drainage. Also, if the friction is theoretically zero in some dimensions, the cartridge is free to wag the tone arm without inhibition and is not restrained around the groove modulation. So this system is not restrained or drained as with a mechanical inerface.
If you have an air bearing platter AND an air bearing tonearm, you have two theoretically frictionless points of isolation mediated by the cartridge interface. This seems like you have one tail (the cartridge) wagging two dogs (platter and tonearm) against two points of isolation without drainage, relying on perfect damping from the two points of isolation. I don't really see how this is ideal even in theory.
Also, on SME website, they cite that material density and stiffness are at least as important as sheer mass. The SME tables create a "mass effect" by using stiff, high density horizontal plates and turntable platter rather than sheer mass.