I don't think there will be voice acting. Having speech bubbles rather than subtitles means that all the visual placements have to take the speech bubbles into consideration. Why would you bother doing that if it's just an alternative to the spoken word?
Additionally, in the types of logic puzzles they've shown (e.g. the arm-wrestling puzzle), you would generally want to be able to review multiple statements simultaneously. This just isn't possible without them being written down (unless maybe you are some kind of super-ginormous genius).
I think it will be like Professor Layton, in that cutscenes may be "voiced" or... "sound-effected" to some degree, but puzzles at least will almost certainly involve the speech bubbles. That would also fit with the style of Annable's animated shorts.