First time I read an argument against all black in design was this 2012 article. Something about it not being a color found in nature. Then, yeah, some people seemed to interpret it as “use gray on gray”.
A few things are truly black, because you don't get to see their color. But "nothing" has actually the color black.
Anyway, even for painting, the fact that they have a paint for it means it happens on nature. It's just very rare, so everybody has to learn the lesson that no, you don't use black paint for that black thing you are seeing.
For computer interfaces the entire thing is obviously bullshit since the beginning. But well, our area has a thing for accepting obvious bullshit.
In the case of characters, especially hanzi/kanji, ones that are hard to recognize/distinguish either evolve (simplify and/or develop disambiguating features) or fall into disuse. Clearly “hard to recognize” is driven by physiological structures that are by definition “natural”
Older scripts, and those descended from other roots, seem to have this property.
Of course in practice Arabic writing is as easy or hard for humans to learn as any other system, so the evolutionary pressure/affordance to get rid of hard parts still applies.
But the theory of early visual system affordance could in principle be studied (though I don’t think fmri is really solid enough yet). Any historical implications of course will always have to remain theoretical.
Writing has technological constraints as well; look at cuneiform vs, say, Pali. Since Arabic evolved after predecessor systems already existed, technological affordances clearly had a, or the, major influence.
Edit: I read your comment again, and just to be clear: your question is really “why is Arabic different”. Latin, Georgian, Aramaic, Brahmic, Chinese, Egyptian Hieroglyphs, Maya pictographs, all have roughly the same form factor.