I think the “open source vs commercial” framing can be adjusted slightly: Proprietary UI products are great and FOSS CLI software are great too.
The difference is that businesses require scaling, parallelism, distributed stake holding always, and that many parts of non-code product developments scale, while code authoring don’t/must not be.
That causes open source UI to be just under-resourced development compared to proprietary, and proprietary code to be just failed development because of over-distributed decision making.
“Under-resourced” is a good way to put it. There’s often a lack of UX design and thus a lack of polish – since there’s often no money to be made, there’s little to no incentive to bring it up from “works for me” to “works for everybody seamlessly“.
(I’ll spare us all the Pareto 80%/20% theory here, although I am pretty sure that it applies here.)
The upside is that open source or free software features fewer (and often no) antipatterns modern proprietary software suffers from.
There’s also quite a scorn in parts of the open source world for UX and design in general.
Attempts at making software more approachable are accused of dumbing down or prioritising form over function. There’s an attitude that it’s not ‘real work’ if you don’t have to memorise tonnes of inscrutable commands or spend ages tinkering with config files.
There are many big ego dudes that think they know what good UX is , they never do any usability tests. From my memory only Unity did real usability tests and for GNOME many-many years ago some student attempted to do something.
So IMO a dude that read some UX book is not entitled to "know better" what actual users need, especially people that do UX design for an app that they don't even use , so they apply dogmatic principles , remove features that were added because users asked for them( this happened at my work in the past too, designrers removed stuff , stuff that users asked for and developers spent a lot of effort to implement it , ex deleting more items at a time, because they think that repeting same 3 click operation over and over again is enoug.
UX should be done by competent people that actually used the programs or even better that test their designs with actual users.
The difference is that businesses require scaling, parallelism, distributed stake holding always, and that many parts of non-code product developments scale, while code authoring don’t/must not be.
That causes open source UI to be just under-resourced development compared to proprietary, and proprietary code to be just failed development because of over-distributed decision making.