One assumes a wall that provides 'a perfect rebound' is one that does not flex and absorb energy, but also does not move (so does not obey conservation of momentum).

Why do we expect an increase in frequency to increase the eddy currents? That would lessen the skin depth in the metal, and the effective resistivity of the metal would go up.

'Step pulley' usually means a drum with two different diameters carved or cast in a single chunk of material. Those diameters rotate together, or not at all.

@Acccumulation No angle-of-incidence argument is available for (for instance) the gravitational lensing, but the Huygens' principle still applies. It's not the total thickness (a global property) but the time-of-flight between points on each side of the interface that determines the light path (and obeys Snell's law). Huygens principle is more general than Snell's refraction formula.

@Acccumulation: In the Huygens picture, it's always a local minimum time that determines refraction. The Fresnel lens has lots of disparate localities, which fail to be smooth/well-behaved at the breaks. Most of the light doesn't hit those narrow breaks, of course.

Radiance being 'watts per square meter per steradian' from a flat surface, means that 'per steradian' part calls for the multiplication by cos theta to get watts per square meter. It's the object shape effect for a flat surface (wouldn't happen for a sphere).

@D.Halsey It is only to a two-dimensional being that the fold has no curvature; such a being has a distance metric that only follows paths in two dimensions. So, the presumption is a non-Euclidean metric space that is not accessible to our observations.