I like how Techno-Speech sneakily released the beta the day before Crypton's announcement.
@Cluemily has IA English and says she has problems with the notes being auto-split (it splits them into unpronounceable words).
New icon lacks the arrows on the upper right.
The default 3 CeVIO voices don't need to be reinstalled (Takahashi, Satou, Tsudumi), but the new ones do. Takahashi sounds the same to me, maybe in the deep learning engine update he will sound a bit different.
I want to share some screenshots/opinions, but I don't have time right now (have to go to work soon).