The thing about open source projects is, they tend to be slightly generic things like "file format," "widget library," or "machine learning algorithm" that most companies see as the kind of thing they can typically get for free from open source. (They're not wrong there - the fact that we're talking about my open source contributions here makes it almost tautological.) They also tend to involve the sorts of generic programming skills that, by virtue of being stuff that we all know at least somewhat well, aren't all that valuable. Supply and demand works with job skills, too.
By contrast, my commercial work tends to involve a lot of specialized knowledge that's quite uncommon, and also tends to produce the kinds of software that people are quite happy to pay a lot of money for.
Counterpoint - I have. Many times.
I created some Perl libraries (most notable <a href="https://metacpan.org/dist/DateTime">DateTime</a> and <a href="https://metacpan.org/dist/Log-Dispatch">Log-Dispatch</a>) that were used by a huge percentage of companies using Perl for applications (as opposed to simple scripts or sysadmin). This often came up in interviews, and I'm sure it helped me get more and better offers.
Nowadays, as fewer companies are using Perl for applications, it's probably less helpful, but it doesn't hurt.
There was a company I know of that was advertising in their job postings that the guy that wrote Perl's DBI once worked for them or was their CTO or whatever. That was about 6 years ago, so not that far off.
But, anyway, yeah - I think that this was different 20 years ago, back when open source hadn't quite completely disrupted the software industry and companies didn't yet understand the economics of open source.
They gave a crap about mine, I believe for two reasons: first, I put it in the middle of my CV, at the same level as my regular job. This basically forces everyone to talk about it. Second, Monocypher is a crypto library, and cryptography tends impress people.
It's not valuable in terms of potential to market/usefulness/popularity, rather, because I can see very clearly how a person designs/writes code, and how they solve problems, when they have plenty of time; I can also see other non-technical aspects, like how they structure their git history (which, to me, has lot of signals).
People (rightfully) dislikes to show their capacity under pressure; open source is the way to show it without pressure :)
