I’m pretty convinced that if I could have my engineers work 4-5 times 4-5 hours blocks of deep, focused work per week - meaning work that is at the very limit of their cognitive and programming abilities, and do another 8-10 hours of the necessary corporate overhead, that we’d collectively get at least twice as much done per year.
One difficulty is that a lot of the work that needs doing isn’t anywhere near anyone’s cognitive limit, but even with that set aside, it seems impossible to pivot to such a world with teams beyond a very small total company size.