"So Windows 10 is keyboard and mouse first, and touch second. It's just a really awful second, and even though Windows 10 has been on the market for two years now, it remains botched."
And even with that being true, it's still crap on keyboard and mouse systems too. Microsoft has only succeeded in making an OS that sucks equally on multiple platforms. If they make it more touch-friendly, it's Windows 8, which was soundly rejected on regular PCs; if they make it more mouse friendly, it's Windows 7, which is worse than 10 with touch (which is one reason it remains the gold standard for PC user interface design).
It simply isn't a good idea. Superficially, it seems like it would be; a large tablet looks enough like a laptop display that it's tempting to think of it actually being one on a part-time basis, with software seamlessly bridging the gap, but it's an illusion that goes away when you actually try to create the OS in question (or the device, for that matter... a convertible tablet is very top-heavy, with the added weight of the touchscreen and the electronics in the screen rather than in the base, which is built to be as thin as possible for portability). That's not to say that such a device can't be useful... just that it's not as good at being a laptop as a genuine laptop.
Windows 10 is about as good an example as it is possible to imagine in terms of demonstrating the difficulties in trying to fit one UI to different devices with different input regimes. If there is a way to do it reasonably well, it could only be including two complete, redundant UIs, but while that would accomplish Microsoft's goal AND please users of dedicated mobile and dedicated traditional devices, it doesn't seem so "gee whiz" cool. It's inelegant to simply include two complete UIs; it's too easy, like it's cheating. It's the equivalent of a web site having a fully-formed desktop version and a fully-formed mobile version and serving them up based on useragent, which is currently considered bad practice.
The buzzword (or phrase) now is "responsive design," which means the site would dynamically configure itself for whatever device the person is using without useragent sniffing to serve up one of two pre-formed sites, but instead by simply querying the screen resolution and other characteristics of the user's device and using that to dynamically configure the site. (Whether that actually works any better than Windows 10 is a completely separate question... I for one get very annoyed by the giant, content-sparse sites I see on my desktop PC more and more these days). For MS to eschew all of that trendiness and just cheap out by including two full UIs... well, that's just not cool, and MS is all about following trends and thinking outside of the box and shifting paradigms and other, more current examples of meaningless corporate-speak (chief among such terms being "the cloud").
Apple figured it out without having to try it first; Mark Shuttleworth figured it out after having tried it. Now we wait for GNOME and Microsoft to see what is increasingly obvious.