> What worked for printed text has no bearing for what works on an illuminated screen.

I find this hard to believe because in both cases what we actually perceive patterns in the light that enters our eyes. It would seem to follow that only aspects of the reading surface that cause difference in the light patterns (color, intensity, phase, coherence) and the time evolution of those patterns can affect matter when it comes to how well something works for reading.

If we determine, whether from printed text on paper, inked text in tattoos, skywriting in smoke, or anything else that certain relationships that some kinds or combinations of the aforementioned aspects of light matter or don't matter, that should remain true when we are reading from screens or from any other existing or yet to be developed form of reading device.