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.