Modern physics has actually done it quite well. This is because the core of many physics concepts revolve around general principles which can be taught directly

Modern physics has actually done it quite well. This is because the core of many physics concepts revolve around general principles which can be taught directly or by example. A modern undergraduate education in classical mechanics teaches concepts around symmetry and energy that generalize to other areas in physics (for instance, the notion of a potential well giving rise to bound states reappears several times in different problem domains). A modern undergraduate optics education generalizes enough that students should readily understand concepts like evanescent waves and acousto-optical modulation. It's only when one moves away from these principles to something more subtle or less well-understood that the education becomes hairier. But as these are further characterized, compression again becomes possible. Landau & Lifshitz, for example, attempts to do this at a graduate level. Many concepts they discuss are increasingly available to the advanced undergraduate due to better compression and better physics principles / pedagogy.