Even if "only" 2.5% of desktops and laptops use it, that is not a small number by any stretch of the imagination -- that's like 30-50 million people. For comparison, ChromeOS has a market share of about 2.4%, and that's with a fairly concerted effort by Google to drive adoption.
Meanwhile, the Stack Overflow Developer Survey showed a massive jump from 25% of respondents using Linux last year to 40% using it this year. Even if you assume some sampling bias, that's still a huge swing.
Then you have major manufacturers like Dell shipping laptops with Ubuntu, Proton and Wine getting better and better at running Windows games and apps, projects like WSL and the Steam Deck giving more people the opportunity to try Linux out, etc.
Even if there are still flaws and barriers to adoption, those barriers are diminishing and momentum is clearly building in Linux's favor. It will only get easier to run Linux going forward.
It probably won't overtake Windows or become a huge competitor (at least, not in the foreseeable future), but I think in the near future it will become more of a "mainstream niche" like OS X was in the 2000s. Most people won't run it, but they'll probably know someone who does and it will be at least conceivable that they could run it themselves.
Google is playing the long game on ChromeOS though, by getting it into schools— those are kids who are going to hit adulthood in 10-15 years with their primary (and maybe even only) computing experiences having been iOS devices and Chromebooks.
Which is all well and good for Google, but that's just not the kind of time horizon that open source projects can work on.
Yeah. There's two camps of the "year of the Linux desktop" topic. "I want Linux to be the majority of usage share" and "I want Linux to be good enough for me to use." The 2nd goal has been accomplished for many potential users since the mid-2000s, and only increasing especially since the web has nearly killed off native application development. The first goal is one I'm not interested in, myself.
Desktops themselves are increasingly becoming 'irrelevant' for people who aren't specialists or using them for office-productivity tasks.
Judging from what I see from my family and people I meet, many homes no longer have any kind of desktop or laptop computers, and rely exclusively on mobile devices of various types.
The era of the 'home computer' will probably turn out to be a "brief" window between the early 80s and this decade. Some people who need specific functionality might have a laptop. But even there, tablets or other restricted mobile-type devices are making inroads.
