I wrote a longer post, but I'll keep it short: you can't compare 100 small res cameras + an inference engine that combines the samples together (our brain with low res foveal samples) on the same metrics you'd judge a single camera with 100x more res. They simply see, resolve and process things differently.
To stick with your example: the pilots that "see" aircraft at 4 nautical miles don't "see" them because their foveas have super human pixel density. But because high-contrast very low res foveal samples + contextual inference + motion + time, give more time to the brain to infer the presence of the aircraft. But if you had the ability to stop the motion and image the very same aircraft would "disappear".
To stick with your example: the pilots that "see" aircraft at 4 nautical miles don't "see" them because their foveas have super human pixel density. But because high-contrast very low res foveal samples + contextual inference + motion + time, give more time to the brain to infer the presence of the aircraft. But if you had the ability to stop the motion and image the very same aircraft would "disappear".