Interesting ideas but I think his physics examples aren't all that great. Last time I looked, Euclidean geometry matched the world as I know it. The world that I can see and touch is Euclidean for all practical purposes, so I think it makes sense to call Euclidean geometry just "geometry" and to teach that as the foundation. Additionally, I remember in intro physics I was well aware that frictionless surfaces, massless pulleys, and pure vacuums didn't exist. The point is that we can often assume they do and still get reasonably accurate results.
I'm a mechanical engineer rather than a pure scientist, so that probably has alot to do with my opinion.
I'm a mechanical engineer rather than a pure scientist, so that probably has alot to do with my opinion.