Has anyone done a rigorous experiment, even on just themselves, with f.lux or redshift? Sleep-monitoring wristband data + "did I see a screen, a red-shifted sc

Has anyone done a rigorous experiment, even on just themselves, with f.lux or redshift? Sleep-monitoring wristband data + "did I see a screen, a red-shifted screen, or no screen" would be really interesting to me.