Maybe it's happening the other way around, with Microsoft adding a Linux kernel in Windows for Wsl2, the final hybrid OS may be neither Linux nor Windows. Windows may never be open source but it will eventually have enough open source baked into it it may no longer matter.
Yes, that’s where I was trying to take this. Open source parts can always (GPL3 excluded) be pulled into other environments, so it is basically impossible for a Linux environment to be “better” than an alternative, unless the “better” you seek is software freedom.
I’m guessing that the 10 years you spent on other Unix led, at least in part, to subsequently choosing Linux. That’s quite a niche experience, one mainly encountered by tech enthusiasts or professionals, and cannot be a counter to the argument that the reason desktop Linux is not yet popular is a lack of compelling reason for many people to switch from the popular platforms.
I first switched to desktop Linux in around 2001, after people on technical usenet forums said that it helps you understand computers better. Incidentally, that’s not true: I learned how to use Linux and Unix tools and the frustration of editing XFree86 configs. Anyway, that tech enthusiasm, or commitment to the idea of software freedom, is a common reason for enthusiasts or pros to switch to Linux. But it can’t carry everyone else, and for many situations a Linux desktop can at best be just as good as what they currently have, and therefore not compelling to switch to.
I didn’t fully colour this thought in because the article was already getting too long, but my position is that that is the best that can happen. The bazaar model means that there isn’t the same pressure to have a well-integrated environment of free software tools as there is for an integrated vendor and that therefore any free software “killer app” will be adopted on Windows/Mac/iOS/Android/whatever and not lead to switchers.
