Software usually isn't monolithic. And this software, in particular, is way more complicated than you give it credit for. Consequently, you overlook many variables that would effect your casual testing.
Right? My brother in Christ, I could add a single line to one text file and reddit would be gone from my life except for the fact that people on Lemmy seem to never shut up about it. I keep it around for research, but I could live without that too, if it mattered.
The "cost estimate" for running Wikipedia, which that article just asserts as gospel, is a "a casual 2013 estimate by Erik Möller, its VP of engineering and product development at the time." So a very OLD and uninformed guess by someone who wasnt directly involved in finance. To me, that makes this read like a sensationalized hit piece, not credible journalism.
Source? That doesn't match up with their published numbers, so I'm skeptical. Post up a source if you're going to make such an absurd claim.
Their endowment is around $100MM USD, as of 2021, which is nowhere near enough to cover their operating costs just from interest. Unless they're somehow obtaining 112% interest...