Lets say you have enough spare time and energy to make a jig, a trolley, an improved tool of some kind that helps you hang siding. For the time you spend making it you are behind on your general average siding panels per day. Once you finish your tool you will surpass your average. This is one of many many non-linear holes in your theory.

You may say 'but you can't guarantee I will make a jig, a trolley or an improved tool', at which point I give up trying to argue with someone on the internet and feel slightly deflated that the point was missed

Iirc, there are several of studies that imply that if you have the same developer working 40 vs 32 hours, the difference in output is often negligible, since not all hours worked are the same in terms of effectiveness.

You might not be aware of that effect, since you deny it explicitly when it comes to your own work, but you might view it in a different light if you turn it around:

Would you produce twice as much if you were to work 80 hours? Probably not, right? You'd maybe produce about 1.5 times as much, and the quality of your work would suffer greatly, leading to more work in the future.

The same effect might work in the other direction, up to a certain point.

The 40 hour work week is not god-given, in fact it has been reduced quite a number of times in the past (mostly due to unions, like most working condition improvemens). So I can not understand why so many people act as if it were set in stone.

Yes. Heavily.

Even in cases such as manual work it doesn’t have a directly proportional correlation, since people get tired.