They screwed up execution before, certainly, and in never was a huge fan, but they were at least trying to make what they sincerely thought was a intrinsically good desktop experience until 8.
Windows 8 was when they had the fear of Android and iOS and the Microsoft phone os was failing on its own, so the mission for Windows 8 was to throw the desktop user experience under the bus for the sake of trying to bolster the phone platform, and maybe make PCs that were tablet like. Also seeing Apple and Google succeed with Internet account based access to the devices was a motivation to get people into an online ecosystem that would have the way to indefinite monthly payments and an app store where they could take a cut off all the application vendors' revenue.
The whole saga with the Metro UI is sad to me too, in retrospect I like that some big player was doing something entirely different to Android and iOS.
The touch gestures and animations on Metro UI IMO still are the smoothest and nicest I've seen.
I feel (probably mistakenly) that if they didn't barge the mobile UI into desktops, that it would've benefitted both Windows 8 and Windows Phone. Still have that flat design for the brand consistency but a more sane start menu.
Not to mention that Win8 itself (in my experience) was the best performing Windows for modern PCs, it had a lot of minor optimisations and not as much bloat as Win10. I daily drove it until the support date completely ended for it, but with OpenShell of course.
The literal entirety of the retail market across the planet has a concept called a "loss leader". You take a monetary hit on Product A with the expectation that people will come in for that AND by virtue of the service being readily available, also buy Product B, C, D, etc. I imagine, if done with minimal intrusiveness, Microsoft could easily do the same with Windows.
On paper as a concept it works. I'm not sure I'm savvy enough to understand the actual issues though.
They employ a lot of this strategy (the ads, the 'subscribe to Microsoft 365 today' 'your computer is at risk to ransomware because you aren't paying us for onedrive'). Except the "loss" part. In fact, I think it's rare nowadays to find a "loss" leader, they seem to have settled around at worst "barely profitable" in business. Too many loss leaders had pretty terrible business outcomes, so it seems to be an unpopular thing to expose your business the risk of going negative margin at any significant scale. Like this disaster: https://en.wikipedia.org/wiki/I-Opener
I must be behind. I've been under the impression that pulling people in on a minor loss meant greater margins. Perhaps COVID had a significant adverse impact on that as well.
Even before covid, I think companies got a bit skiddish about actually going negative. Probably still do on little things, but I think Microsoft at least makes Windows pay for itself while also using it as you would use a loss leader.