This makes total sense.

IMO the retina isn't so much "a window into the brain" as a part of the brain that is potentially visible through the lens/cornea/humor (now a standard bit of gear in many ophthalmologists'' offices)

The retina is a neural net, and clearly some processing (edge detection and potentially shape detection) happen right there. Dr. Jerry Lettvin at MIT did some interesting work showing this in 1958-9, published as "What the Frog's Eye Tells the Frog's Brain". Here's a PDF: https://courses.cit.cornell.edu/bionb4240/Reprints/Lettvin_Maturana_McCulloch%20and%20Pitts_1959.pdf

I ran across this paper when I was getting into neural network programming, 20 years post publication, I knew his son (scary-smart software guy), got to talk with Jerry (also brilliant - but in neurobiology) about it over several dinners.