[randcraw]

True, but Q&A makes up a tiny fraction of most lectures I've attended, and attention span can be improved by shortening the duration of each lecture.

However, I'd agree that lack of dynamic peer interaction is a big problem with video-based instruction that hasn't yet been fixed. I've seen attempts to address this like Piazza where students can post questions or answer others', but text-based forums lack graphical or temporal cues that many concept demand. Maybe some sort of multiuser video supplement might help, where the question poser could snip the time mark of a puzzling section from the lecture or an illustration from another video and refer to it in source so others could visualize with a click the point being asked?