Guitarists can usually tell when one of their strings is slightly out of tune and can often tell when all their strings are out of tune yet in tune relatively to each other. Successful guitarists will have a guitar tech off-stage; swapping guitars isn't just for tonal purposes, it's to quickly swap out a slightly out of tune guitar for one that has just been tuned. But it seems unlikely that without perfect pitch that we can trust a non-musicians' assessment that a musician played a song perfectly faithful to the original. Even assuming that this occurred, this is not exactly all that unlikely nor that it rarely occurs, but, instead, it is academic for the experienced performer and trained musician. And for every hour of performance there are many hours of rehersal. The very best performances are built from muscle-memory which will obfuscate how much raw talent there actually is.

I've been a guitarist for 35 years who took an ear training course w/ vocal lessons about 10 years in.

I could probably do better than blind determining what is 432 vs. 440 because there are enough common reference examples out there that I have sense of feel for the difference, and I could tune an A4 cold +- 5 cents. But a third of that? A vanishingly small number of musicians would be able to do that with any reliability without a reference point.

She didn't know how on-point a pitch was because she was listening. It was innate and immediate. It was literally described as the intensity of a color that corresponded to a key.

We're kinda off in the weeds from the point I was making, which was refuting the notion that music could have any analog to reciting a phone book. There is a bit of a connection between memory and processing music in a way, at least for those with this condition, that probably is more subconscious for those without.