Make theorem 4.3.4 (only five regular polyhedra) into scaffolded exercise.

[oscarlevin]

Maybe. This seems to be too complex to include in the main exposition, and the complexity doesn't have much to do with Euler's formula.