The evidence that blue light is bad for sleep is much worse from an experimental standpoint than the evidence that it improves cognition and reaction time, which are obviously rather important considerations in street lights. Consider for instance that placebo glasses which filter the so-called harmful blue light wavelengths have been shown to fully reverse the apparent effect on sleep quality. This is despite them just being clear glass with no appreciable filter effect on visible light.

It is literally true that human mind is suggestible to an unimaginable degree.