I'm really excited about that too! I hope companies try out some experimental voicebanks - even if they don't turn out perfectly realistic, the results could be very interesting and unique.
VocaVerse Network
lIlI
Prism
Having done ai tts one would need to denote the special kinda of singing so you can have control of it. I made a tts of a singer and with out the denotation she would randomly belt. It might be best to make 2 voice banks to have full control over it like vy2s falsetto or appends