Interesting theory, but if that were the case we'd surely see more variability in the other days of the week too. I can't quite understand how power-down events affect this rate anyway? Perhaps the mechanics of the 102 week power-down are different in some way? In any case, if this data is valid, then some days are slightly better for powering down than others, which seems quite strange.

I have assumed linear interpolation between certain time points for the steemreports financial tools, so wanted to see how accurate that was, and this has confused me a bit! ;)