In my proposal, the number of hours a student actually spends directly listening to/interacting with an on-premise educator is much, much lower. The hope is that those would be more efficient hours, to the point of getting better results with >5x less time.

Part of the issue here may be that I found the education system to have net negative impact on my... education. The main value I see in the current system is providing child care, safety, and food security. If there is solid research showing that ~40 hours of the current approach outperforms ~5 hours of a tutoring-style approach in the general population, that would be evidence against my proposal. And that might exist, I just haven't seen it, since this is not exactly a formally vetted idea.