Very much this. I'd recommend using these books as adjunct material. I found them indispensable as an undergrad when I was struggling to shift from a mathematician's rigor-and-proof perspective to a physicist's intuition-and-approximation perspective. However, I don't think I could have come close to passing my QM or E&M courses, even with a mathematical background that was stronger than most of my peers, if I'd only used Feynman to learn the physics.
Is this an example of the kind of reactive, interactive design you're advocating? I don't see how this is helpful: it's dogmatic and doesn't address the actual tradeoffs that might be at play. Why 64 ms? And 32 and 16? What studies are you relying on to suggest that 65 ms is too slow or that 30 ms is too slow for initial feedback? How does it affect user behavior? What is advised when these rules, especially the "full screen of content" cannot be met because of things like external dependencies out of the site's control?
I think we have pretty good evidence in recent years that this is not how things really work. Those filters and algorithmic steering just provides a way to keep people in their bubble. What you’re describing is basically what YouTube was: allow anything legal but use sophisticated software to show people what they’ll respond to positively. The radicalization this led to is well documented at this point.
I’m not really in favor of what CloudFlare’s done here, but it’s really a fact of basic statistics that exposing the mainstream to well-made propaganda is far more dangerous than exposing radicals to the mainstream.
Imagine 1% of non-radicalized people exposed to radicalism become dangerous and 50% of radicalized people are de-radicalized by exposure to the mainstream (laughably optimistic). If you start with 100 million non-radicals and 500 thousand radicals, you have a 125% increase in radicalization.
This is how we now have so many flat earthers. When you have massive reach, it’s easy to grow your numbers.