Perhaps not with pipx itself, but as a symptom of what's wrong with Python (and most software nowadays) there are lots of problems with it: Everything is such huge stacks of stuff built on other stuff built on other stuff built on other stuff... All huge tottering heaps of infrastructure where shit can (and does) go wrong in every layer. It's package managers running in containers calling dynamically-linked libraries running on virtual machines calling package managers running in containers calling dynamically-linked libraries running on virtual machines...
With, say, Turbo Pascal you compiled your code into a .COM file, copied it to a floppy, and people ran it. Just your program. On their OS. No heaps of layers of infrastructure and toolchains to get knotted into a tangle of failing interdependencies. How the fuck can it be so impossible to build stuff like that nowadays?!?
Not that I'm blaming pipx (whatevertheheck that is; judging from the name seems to be a relative of pip's, which I've heard of) in particular, or even Python itself, more than anything else. Almost everything nowadays seems to be that way. I just wrote all this in the hope that it might explain why a lot of us (like for instance your interlocutor saghm in a sibling thread) are rather reluctant to pick up all the New[1] Shiny stuff. It all seems to so inevitably bring with it rat's nests -- colonies of rat's nests! -- of interdependent cruft tangled up into tight balls of mud and misery. You lose the will to even start trying anything out.
___
[1]: And yeah, OK, calling Python "New" might be problematic... If we'd gotten around to picking it up some fifteen years ago it might still have been. And maybe not accrued as many layers of cruft.