Great thread. Hope the new home is all you hoped.
Your results map closely to our experiences using a TriPlanar/UNIverse and Cello's Graham 2.2/UNIverse.
As SirSpeedy first reported and we heard, the Graham unipivots also benefit from very fine, cartridge-specific adjustment of damping fluid around the bearing. It makes perfect sense that a JMW would do the same.
Most TP owners who hang around here have followed my suggestion to ditch the supplied metal AS weight in favor of something much lighter, to apply a very small amount of AS. We use O-rings and I believe VPI started offering O-rings with their AS option shortly after I mentioned that trick here. I should have patented that! ;-)
Without the anti-skate, his voice was centered in 3 dimensions exactly between the 2 speakers. I sensed a kind of blurring with A/S that was cleaning up without it... I suspect that although the A/S gizmo does its job as intended, it also provides a bit of damping to the stylus which simply may be too much in my system. This would explain the improvement using it with the arm not damped.
Exactly right, and easy to understand when you think about it. Just as with VTF, AS applies a constant external force to the stylus/cantilever. This presses the cantilever against the elastic suspension within the cartridge, which of course dampens transient response, limits dynamic extensions and smothers HF's. Just as with VTF, excess AS smothers and dulls the sound. Imaging loses precision because you've lost HF's, which our ears are most sensitive too when determining the direction, size and shape of a sound source. Images restricted to the lower frequency sounds from a source will always sound larger and more diffuse than images which include the higher frequency components.
Great observation picking up the relationship between well damping on a unipivot and AS. I suspect a similar relationship might exist with VTF. Sharing stuff like this is what makes this forum so valuable.