I think the lectures need to be replaced by the videos as a home assignment. Classroom time should be split into two activities - intelligent discussion on the subject of the video, and actual problem solving with other classmates, with the teacher immediately accessible for any questions.
My intention was to invert the role of homework and classwork. That is, that homework would now consist of watching the lecture. Classwork would focus on solving problems. The outcome? Disaster.
I tracked usage of the videos. Around half the class didn't watch a given lecture. Around 1/3 consistently watched the lectures. The effect of this was that over time only around 1/3 of the class was able (and willing) to engage in problem solving and discussions about the material.
I had 2/3 of the students hopelessly lost by the tenth week. How does this compare to the traditional format? When I give the standard lecture/homework assignment type class around 2/3 will pass. I double the number of people passing but I strongly suspect that they end up failing in the next course. (It is quite easy to pass beginning math classes these days. ) My anecdotal experience is that while I get more people to pass by giving lectures each day I end up with a group of people who know the material less well.
My experience tells me that around 1/4 of the students will learn the material from my video lectures because they are motivated and willing to engage me when they are lost. Around 1/4 will stumble through the course and learn something regardless of the format it is given in. The rest are mostly hopeless. They lack motivation, drive, and the ability to work intellectually.
What you describe does work with motivated students and the right teacher. I am not the right teacher for this though. Without motivated students I don't think it works. That's the essence of the problem. Too many students don't want to learn.
In a traditional class a student can hide their ignorance. They don't raise their hand and don't participate. The facade of being studious and understanding the material can be maintained. When you assign lecture watching as the homework and then spend class time solving problems it becomes evident who is not keeping up. It can't be hidden. This causes negative reactions in my experience.
Here's a great lecture by Eric Mazur. It's long but well worth watching. It's called, "Confessions of a Converted Lecturer".
Additionally, watching the video could be an assignment that counts towards a grade. Sure they could sign in, turn on the video, and walk away, but this at least provides additional motivation to do the first two steps, which may be all you need to overcome the activation energy for many students.
Additionally, threats of failure help - I had a course that was considered the filter - about 50% failed every semester. I think I was motivated by that statistic because it ended up being just about the easiest class I had. The motivation was less about not wanting to fail and more about showing excellence where common people fail.
Motivating them to watch is a problem for me. I'm philosophically opposed to giving points for something a well trained monkey could do. I can't bring myself to give points merely for pressing play on a video. I'm slowly coming around though.
So the problem with having videos that can be watched any time as opposed to going to class at a set time sounds like it could be a good idea. But I'm guessing it wouldn't work that well--except, as you suggest, with very motivated students.
Overall I think the problem comes from a lack of a desire to learn. I'm not a motivator and so I think my students suffer from my inability to be a rah, rah type instructor. I don't know what can be done about it though.
When watching almost any tutorial video, I find myself zoning out a lot. Often times I even open up the browser and read something completely different while the videos is still running.
A few weeks ago I had a discussion with a friend who challenged me to convince him his adamant assertion that videos are the inevitable future of all media is wrong. At the time I just felt instinctively video was never going to replace the written word and 2D graphical presentations, but I couldn't make a convincing argument. It only now occurs to me there might actually be scientific data in support of my suspicion and that instructions on video may have an information-to-brain bandwidth similar to a very boring teacher after all.