Edit to answer the first point: free software as a cultural movement grew out of a need to control the hardware a software engineer bought and which came with no source code for the drivers. Friends made along the way was great, and I agree that's lovely, but ultimately that's what free software enables, among other lovely things (like not depending on a manufacturer you disagree with, or another programmer you disagree with).
Forking a project does not duplicate the effort — collaboration requires a lot of effort, and the more collaborators, the more expensive it is (usually even more than 2x). This is why many well-run free software projects have their own BDFLs (Linux, Python... OK, maybe not so benevolent at all times :)).
If the two of us wanted to take a project in a different direction, and insisted on keeping it a single project, we'd have to do more work to reach compromises all the time, neither side ever being fully happy, but doing duplicate work for an imperfect result.
As you are well aware, RoI calculation is not as simple as LoC count, and typing in the code is the fast part. With two projects with clearer missions, willing contributors flock to the one they align better with, which leads to one being more stable, another being quicker to add new features.
And no, getting all of those contributors on a single project would not result in a single project that is both stable and adding new features quickly.
Obviously, no expectation on you to do anything. I just find it surprising you do not see the value in a different way to organize a community compared to what's common, even if what's valuable to you is different.
The escape hatch is the feature that force everyone to be nice.
What if you don't like Windows 11 forcing to make an account? You get a time machine, go back 30 years and start a hobby OS project and hope the best.
What if you don't like Linux? You fork the project and just ignore what Linus publish in his tree. (The hard part is doing a better job than Linus and convincing enough people that you can do it.) So Linus must be nice enough to keep most people happy.
Not _most_ in the sense of a plurality of all users. _most_ in the sense of some subset of contributors. Free software is still a dictatorship of the able. A person who cannot contribute (whether that be code or money to pay someone else to code) something important is still without recourse.
If Linus decided to add a bunch of malware to the kernel, my mom would still be unable to fix that.
FOSS is still better than proprietary, but it's not perfect either.
You are right, but your mom would easily benefit from someone else who would be able to fix that: free software relies on the fact that someone "able" is at least not looking to hurt you, which is, for a popular enough software, a given.
For software that is not popular enough, you are right that in practice, your mom wouldn't be able to do anything other than stop using the software.
But of course FOSS is not perfect, it's just a way to empower your users: nothing more, nothing less.
Forking a project duplicates effort and requires much more work than just allowing collaboration.
Again, though, I have no particular desire to do anything here. I don't even use my vacuum.