[mattkrause]

Obviously, the last rocks don't unsmash themselves. However, if someone plans to work 60 hours a week, they might work 9% slower--perhaps even unconsciously--thereby causing their output to be be the same as if they were only supposed to work 55 hours.

This is borne out in data from a British bomb factory during WWI. https://onlinelibrary.wiley.com/doi/epdf/10.1111/ecoj.12166

Figures 1 and 2 (page 2060 and 2061) show that, across four cohorts doing different tasks, output plateaus at about 48 hours/week. Indeed, output from 70 hour week (10 hr/day x 7 days) was slightly lower than a 48 hour week (8 hr/day, with Sunday off). These workers were pretty motivated by the circumstances and doing skilled but not particularly creative work, so I suspect this is likely an upper bound.

There's lots of interesting data about working conditions from the the Health of Munition Workers Committee.