If you make that argument about the state of software in general, I’d agree to an extent in the sense that it should be more prioritized. But I don’t see how that applies to open source in particular?

In those aspects proprietary software is just as bad, if not even worse. The difference is simply that the default choice of software for most tasks is a proprietary software. They can have a shit ton of unusable and confusing mess, even intentional dark patterns, but users will adapt.