Apparently those mm waves liked that plastic lens just fine.

In fact, that would work fine at 10GHz, at 1GHz, and even at 1MHz,
though the amount of material you'd have to use for the lens gets
prohibitive at lower frequencies. It's all engineering tradeoffs. I
know that "geodesic" lenses are used in some radar systems; the idea
is that you have the signal travel a longer path (through a curved
waveguide structure) in the center of the antenna/feed than it does
toward the edges, just as in a convex lens the light in the center of
the beam is slowed for a greater distance (and therefore retarded
more) than the light at the outer edges.

I expect the boundary between "optics" and "electronics" will blur
even more than it is already as both electronics and optical
technologies continue to advance.

Cheers,
Tom

Hi Tom, we've used plastic lensing since at least the late 60's for focusing
mundane 4-12 GHz radio waves. Dielectric refraction was used back then to
extract additional gain from dish antennas by allowing more even
illumination of the dish without illuminating the area around the dish.
Harris radio had a patent on it.

W4ZCB