To add some detail: a common minimum quality criterion for optical surfaces is λ/4 (divergence from an ideal surface), where λ is the wavelength used. For visib

To add some detail: a common minimum quality criterion for optical surfaces is λ/4 (divergence from an ideal surface), where λ is the wavelength used. For visible light this gives about 100 nm (though large professional telescopes are "figured" to more like λ/20). Any alleged glass creep visible in window panes over a few hundred years would quickly ruin any astronomical mirror (which definitely does not happen).