The density of damage to surrounding panels suggests that random distribution doesn't explain zero hits to the undamaged. Different strength is a possibility, and replacement of degraded panels could explain why different.
Arrival of photons, rain drops, hail, should follow Poisson distribution. If damaged panels have average of 25 hits, then standard deviation is square root of 25, or 5. To get only 5 hits not 25 would be 4 standard deviations. Zero hits would be 5 standard deviations. Not gonna happen, especially to the density of undamaged panels we see.
Non-tempered glass in thin-film panels (due to processing temperature) would be 1/4 the strength, could take 1/2 the velocity of impact. But these panels are likely single or poly crystalline. Later batch of glass was stronger? maybe thicker, but if anything I'd expect newer ones to be thinner.