Knowledge work does not have 1-1 correspondence between time spent and productivity. Things get VERY non-linear, to the point that more than 50 hours of real knowledge work a week is often LESS productive than 40 hours.