Much of what I've learnt about this hobby has largely come from the writing of two men - Peter Aczel and Alan Shaw.
The sheer depth of research, testing and innovation that goes into a Harbeth loudspeaker, particularly the design of its BBC style thin-walled cabinet, would suggest that a Harbeth loudspeaker might be the one place where speaker isolation would have no sonic benefit.
After all, haven't all the panels, the joints, the assembly, the way the baffle is attached to the frame, the way the drivers are attached to the baffle in a Harbeth all been designed in such a way to render cabinet resonances below the threshold of audibility?
Alan Shaw has previously indicated that he doesn't even believe that purpose built stands are necessary. Didn't he once test one of his designs as it was casually stacked on top of a pile of telephone directories?
So if some form of isolation is having a measurable effect, then perhaps someone should tell Alan?
Could it be that even the much vaunted relatively low mass Harbeth cabinet is not as free from internal resonances (or ringing as MC likes to put it) as was previously thought?
I would tend to think anything mechanical must be affected by its relationship to the surface it's placed upon. Therefore it seems obvious that this must, to some extent at least, (depending upon the speaker and its location) also apply to loudspeaker performance.
However it would still be interesting to hear what experienced loudspeaker designers think of this issue.