I'm sure the CA team is working on this however it would be great if we had an option to drag and drop the visemes in to the timeline for better precision (if there is already a method please let me know).
I'm currently using text layers with letters in photoshop to aid me during playback.
Thanks - and yes, we are working on something much better for viseme editing. It won't be in the upcoming preview 4, but we're working on it - and it will make lip sync much more user friendly in general.
Currently "split lip sync into visemes" is the best way, but it can get cumbersome. See https://youtu.be/AkggrBwI-hg?t=2m48s for more info about this.