Don't want to be pedant here, but your calculation does approximate quite a lot:
Degrees in one inch - arctan(1/24)2 = 4.77
"Data points" in eye for one inch, 12" away: 4.7750*2 = 477 (magic number cited in article)
Why the last multiplication by 2? The eye doesn't see pixels, it sees color difference, which means that the limit (not the minimum) is when two of those pixels cover the same "cycle" -> 1 color change in the cycle (in any case, that's how I understood the article).
Degrees in one inch - arctan(1/24)2 = 4.77 "Data points" in eye for one inch, 12" away: 4.7750*2 = 477 (magic number cited in article)
Why the last multiplication by 2? The eye doesn't see pixels, it sees color difference, which means that the limit (not the minimum) is when two of those pixels cover the same "cycle" -> 1 color change in the cycle (in any case, that's how I understood the article).