Physics help...sort of

Uh... the only tricky thing I can think about this is that the delay per century is compounded. Typically the formula for this is F = P(1 + i)^n, where P is the current duration of a day, i is the % time increase per century at time 0, and n is the number of centuries.

Then there's the energy equation stuff, but I don't see how that would be different.

Frankly that's stupid imo. I think it's just 20 ms. But we could be wrong.