One thing I haven't actually heard is an A/B of the Supernova patch sets played on the Ultra versus the original. I've seen one guy (subjectively) state that his Supernova sounds warmer, but he said that based on the original patch banks of the Ultranova (which are horrible) and unscientific tests (the Supernova banks had not been released yet). I'd love to hear a valid (and properly conducted) test of same patches on both. So far I'm inclined to believe it's another case of "placebo effect" induced by all the lovely knobs on the Supernova
I see (or I should say hear) that a lot in the battle of hardware vs. software synths. A perception of warmth that perhaps exists when listening to the raw signal but gets lost by the time it's placed in a digital audio file and listened to on an MP3 player, over soundcloud etc. Seems a little like getting caught up on a requirement of "pure analog", then ultimately putting the music into a digital format

Ironic to say the least.
I can believe the SN is capable of sounding better overall because it was 8-part multi-timbral I believe? (Thus could get some layered patch sounds that the UN could not) But what I'd be interested in is how it sounds in a direct comparison on same single patch. Supposedly these SN patch banks for the UN are direct ports, so it should be easy enough to do for those that have both synths -- hopefully a youtube video or similar will surface soon.