With the caveat that I am comparing different speakers in different rooms with different equipment at different times (which is pretty hard), I would characterize things as follows:
1. The Ulimates have the highest resolution and best micro-dynamics of these three speakers. The mbls are the lowest resolution in this group. The M5s are exceptional on micro-dynamics, but to me fell short of the Ulimates. D/A conversion may have played a role here as well.
2. The overall tonal balance of the Ultimates seems the most neutral of the three systems I heard. The mbl system I have is a little darker than neutral and JV's M5 system seemed slightly less present in the lower midrange than neutral.
3. The mbls create a soundstage recording after recording that sounds un-hi-fi-like (more like live music) -- they really get the image off the speakers. The mbls don't do pinpoint imaging the way some speakers can, but then live music often doesn't have this attribute either. The Ultimates image well, but the size of the speaker makes it harder to suspend disbelief (i.e. to think that the sound isn't coming from the speakers).
4. All three speakers deliver a package that makes you initially think "this is the most realistic sounding reproduced music I've ever heard". Over time you hear reproduced music artifacts, but there is a coherence to the choices made by the designer of each speaker that is musically consonant. This idea of musical consonance has been distorted by some readers in the past to mean something unlike live music. But it is the opposite of that. Live music is clearly "of a piece". Speakers inevitably have distortions, but there can be distortions that tear at or shred the fabric of the music and those that don't. To some degree, what merely pulls and what tears will be in the brain of the listener, but I could see how each of these speakers could appeal to certain brains as musically coherent.