I'm a little curious of this myself
I think that mostly vorbis is sounding really pretty decent right now, and that pre-echo is by far still the largest issue to be worked out. I'm sort of taking a wait-and-see approach to see how this can be addressed before start to experiment a little bit more with the codec. I've provided a good deal of samples which exhibited post-echo issues which I understand have mostly been fixed now (haven't really tested this yet though I should. I believe ff123 reported that the problems he heard in death2.wav were fixed though). They can still be found here for those interested:
http://www.animus-facticius.org/vorbis/
(Files are encoded with LPAC which can be found here:
http://www-ft.ee.tu-berlin.de/~liebchen/lpac.html)
At least now, I believe the clip which may exhibit the biggest problem is nuane. There is a sort of crunchy, rumbling type sound in the background of this clip which is amplified quite a bit when encoded with vorbis. Garf attempted to tune this problem out, but even with his vbr mode it was still an issue. Surprisingly enough an older build of vorbis encoded this file better at 128kbps than a recent one with Garf's vbr mode at near 180kbps. Anyway, Monty knows about it though so we'll see what happens
Btw, most of those clips are a little extreme in nature, but they happen to come from a genre of music that I listen to quite often even if other people find them rather artificial. I do realize they may not be representative of the majority of music out there, but theres no reason to not try to achieve perfection across the board