So, extending that rule (approximately):

All the actual work on Earth is performed by sqrt( 3,630,000,000 )[1] or:

~60,000 workers

[1] https://data.worldbank.org/indicator/SL.TLF.TOTL.IN

"Based on choosing the current estimates of Labor force, total"

You are missing the implication in the equation that smaller projects/teams are more efficient.

Gotcha, Singleton sole proprietorships it is. Down with complexity, break up every business! /s

You're making a fine argument against optimizing for one variable. But that doesn't discredit the "law".

Half of all of the work - not all of all of the work.

Is everyone on earth working on the same project?

"law" is an incredible term used for "an observation a physicist made about the author citations of academic papers at one point in time", especially when you try to extend that to software development, and realize that there's other competing theories with supposedly better fits. I have not independently re-run the analysis myself, but lotka's law claims to be better an in general these are all special cases of zipf's laws, which is admittedly where I personally first heard this concept.

...which is probably why you only see this stuff regurgitated in blog posts and right-wing Malcolm gladwell (Jordan Peterson is quoted as the source in one of your cited blog posts).

Either way, I'd be highly, highly suspect of parroting Price's "law" as a fact.

(I get stuff like Conway's law or Moore or Murphy are all also cited as laws, and I don't like that terminology either. "Conjecture" would fit so much better, save for Murphy's)

Even if the law were true regarding authorship, and applied to software, that still wouldn't show that the "valuable contributions" are only made by virtue of a small set of excellent contributors -- see "Matthew effect"

https://en.m.wikipedia.org/wiki/Matthew_effect