I still recall back in the day before the OSI decided to politicise (and perhaps even value-extract) the term "open source", there were still a heaps of people doing groundbreakingly useful coding and giving it all away.
Linux is perhaps the posterchild of that, but Apache was just some people patching NCSA httpd because they needed new features. Perl was Larry Wall writing a better language for generating reports. PHP was Rasmus Lerdorf's project to make it easier to build his personal website and make it "dynamic" and able to connect to databases. The vast majority of the commands and utilities in your Linux distribution were written by the GNU project and many of those by Stallmann himself.
Out of those (and many other examples most greybeards here could mention), probably only GNU/Stallmann were motivated by "lofty ideals", pretty much everyone else was doing what was called "scratching their own itch". The needed features or software that didn't exist, so they wrote it.
License-wise, Apache could only exist because NCSA httpd allowed it - it was explicitly released into the public domain (and later GPL-ed I think?). Perl was dual licensed, you could choose GPL or Perl's own "Artistic License" which was much more free, allowing you to sell stuff incorporating it so long as you still credited the original authors. Linux was originally licensed with something Linus wrote without lawyering up, which was basically "the source code must remain available" and "no money can change hands" because he was annoyed at having had to pay for Minix, but shortly after he changed his mind abut the money aspect, and chose to use GPL instead.
While from my perspective, Stallmann was the "lofty ideals" guy, and pretty much everyone else were pragmatists who were "just getting shit done", and sharing stuff so other people could get shit done better too.
It was Eric Raymond and Bruce Perens who hijacked the already in-use term "open source", and claimed to have invented it and went on to build a business-friendly brand and "industry" around it - and as you say those "industry people" will _always_ jump in with their revisionist history to "correct" you when you say otherwise.
So to me, OSI-defined "open source" really can be wage theft. It isn't always, but it's intentionally set up so it can be. (Probably not originally with malicious intent, I suspect back then many of the people involved were perfectly happy with the reputational payoff in lieu of monetary reward.) The existence of "Open Source Companies" like Mongo, Elastic, Hashicorp et al, and BSL-style licenses - seem to me to be closer in spirit to Linus's original license than anything the OSI approves. And I feel they are the result of a new (relative to the '90s) trend of people writing open source software for the sake to writing the software, instead of the old-school approach of people writing software to get their actual jobs done better or to scratch their own itch. Linus and Larry and Rasmus and Brian didn't write/release their software so they could become Founder/CEO of Linux PTY LTD or Perl Corp or PHP Inc or Apache GmbH. They did it because they wanted an OS they didn't have to pay for, or a better way to generate reports, or a better way to make websites, or a better way to serve the HotWired website.