I feel that you do not understand the magnitude of difficulty behind what you are attempting. You seem to think it's going to be a breeze; that you can solve all the problems that the MOOCs around haven't been able to solve. And you think you can do this because... well for no particular reason as far as I can tell. You haven't proposed any idea that is distinct from what is already out there and is evidently not working. E.g. Udacity.
> adding content based on the most common questions, and providing a low-friction way for students to ask questions about the content
My criticism was about individuals asking questions. "the most common questions" are not the best ones, and don't help those students aspiring to learn to a higher level, who tend to ask better and more personalised questions. In fact the better part of science is asking the right questions! It is a talent to do so. What you will receive is a sampling of the lowest common denominator. And most of those questions deserve a more tactful version of "RTFM". This is my experience from tutoring-- students mostly ask questions that were answered in the lectures, because they've been procrastinating and not watching them. Especially a problem when the students don't commit themselves to attending class in person. In fact the best low-friction way for students to ask questions is... to be present during the lecture and put their hand up.
> Frankly, outperforming random profs at random schools is a pretty low bar. I've already done this just by tutoring.
What you are proposing is nothing like 1-on-1 tutoring. Feedback is part of this process. Try asking a "medium" to tell you about your relative as you hold shut lips and a poker face. Similar to a professor trying to give a lecture to a room full of mice.
> It's rare that a live professor adds much value for me vs. some of the best professors online.
In that case you aren't asking the right questions, or you aren't going to a very good university. While my experience of in-person classes was sometimes astoundingly poor, I would never exchange them for an online lecture series. You can talk with a professor and learn from their expert way of thinking-- not just the content itself but the way they think about problems. You learn through interactive dialogue-- people in the real world call this "conversation". In addition, the collegiate atmosphere of study among the pupils is in my opinion an extremely important component in students keeping up with the content. Only the most disciplined people can sit down and complete all exercises from a MOOC without interaction with others.
> the real value-add I've had from professors is when they've done the content-development work to put together organized course materials and/or well-thought-out problem sets well-curated suggested readings.
I agree, being organised does help. This applies to many things in life. Many professors at universities are not organised because of the pressures put upon them by the university administration, and other such factors. Also the pressure to push through any mediocre students to pump the university's finances and public image.
You must understand that, while it isn't impossible that you could create such a platform that is as successful as you want it to be, the a priori knowledge around previous similar ventures tells us something about how that might turn out. If you believe your idea is different, then how? You have not identified any key differences between your approach and those that have failed...