The brain really is quite a machine. I've personally had a retinal tear lasered. It's well within my peripheral vision, and the lasering of course did more damage (but prevents it from spreading). How much of this can I see? Nothing! My peripheral vision appears continuous. Probably I'd miss a motion event only visible to that eye only in that particular zone. Not to mention the enormous number of "floaters" one gets especially by my age (58). Sometimes you see them but for the most part the brain just filters them out.

Where this becomes relevant is when you consider depixellation. True blur can't be undone, but pixellation without appropriate antialiasing filtering...

https://www.youtube.com/watch?v=acKYYwcxpGk

So if your 30x30 camera has sharp square pixels with no antialiasing filter in front of the sensor, I'll bet the brain would soon learn to "run that depixellation algorithm" and just by natural motion of the camera, learn to recognize finer detail. Of course that still means training the brain to recognize 900 electrodes, which is beyond the current state of the art (but 16x16 pixels aren't and the same principle can apply there).

It would be interesting to see how far you could push that. I bet just two scanlines side-by-side would be enough for complete imaging. Maybe even just one, but that would require a lot more pre-processing and much finer control over the angle of movement. Congrats on the positive outcome of that surgery, that must have been pretty scary.