Your Python example is undeniably progress because Python is still in wide use, a better example would be if you improved the performance of Dylan (or some other language that has long since fallen out of use). For your flooring example, imagine if you were contracted to work on someone's house and went, "Instead of using fresh floorboards, we decided to install floorboards that have been sitting out in the rain for 40 years because we enjoyed the challenge!"
I've seen enough people make comments on how contributing to free software is a fight for the future of political and social freedom, since those who control technology will have a disproportionate impact on the systems of power and I find that hard to square with the attitude that free software no big deal and we should just have fun.
If it seems that way, it's because a given online discussion isn't a random sample of people's beliefs. For example, to someone reading Reddit or Slashdot (or HN sometimes, unfortunately), it's easy to jump to the conclusion that everything sucks and everyone hates everything. Post an article about Python, and all the comments will be about how Python is terrible and everyone should use Ruby (or whatever). Post an article about Ruby, and all the comments will be about how Ruby is terrible and everyone should use Python (or whatever). Read the same site for long enough and it will start to look like a mass of contradictions- "wait, everyone here said Python sucked before and I should use Ruby, but now they're all saying Ruby sucks and I should use Python?!?" But really it's just different people showing up to disparage whatever the topic of discussion is. I think it's just the nature of the medium that makes it really hard to gauge consensus within a community.
I've invested a lot of my personal time in creating the fastest befunge implementations I know of. I intend to spend even more time eventually when I get to having it JIT to assembly. I find it fun, I don't get paid for it, & I have a modest salary (under 40k/yr)
It _did_ end up having positive outcomes, where I was JITing befunge code to python bytecode, & it got me thinking about the bytecode's performance, so to optimize my befunge implementation I went a level deeper & optimized the python implementation to wordcode instead of bytecode, giving anyone who uses Python 3.6 a 1% improvement
But maybe working on improving the performance of Python by 1% is frivolous, since Python's just some big codebase whose only use is in silly things like my befunge interpreter..
That said, I agree with you somewhat on that tech industry rant. I mention my salary's modest, but it was less while I was working as general labor in flooring, which was way more than when I was vending icecream off a bike (which was much more than when they'd have me put together coolant packs for 25$ in a 6 hour day)
You're directing it at the wrong thing. Writing a blog about some hobby project isn't what's wrong with the world