We did not manually muck with control panels per station. I could not care less what the settings UIs for home installs of Windows are.
And that's the real truth, power users and these small fry IT install bases are not major markets. The most prominent markets are OEM installs for non-technical home users (who hate complex panels and love the push-button UIs), and large 3/4/5 digit enterprise installs that do not give a damn about this stuff because they're using MDT (or whatever) to deploy images to all their stations and not playing with this per-station configuration stuff.
We all go through that stage, it's natural.
I once installed Linux on all my siblings' desktops because I had convinced my parents it would lead to "less viruses" (Fedora 13, I think, GNOME 2, those were the days). The nightmare of providing support for that over the rest of my high school years provided me with a powerful and important lesson.
You have few reasons to stick with Windows, you are a very, very tiny droplet in an ocean of users with different needs than you. Judging what such markets will do based on your personal needs is unwise.
And honestly, breaking Printers is a thing that everyone is going to care about, especially people in a workplace.
You're right that the home market is shrinking though, the OEM install market is down billions in revenue and has been dropping by double digit percentages each fiscal year, but you're wrong in your assumption that the power user market is worth anything. Power users, represented by direct license installs, are a rounding error of Windows revenue. There's nothing worth fighting for there.
Convincing Mom that she should buy a Windows laptop instead of an Apple laptop is still worth ~$750M/Q, and Mom doesn't ever want to see the old Windows panels, she loves the "Add Printer" button. Convincing Timmy Q Hacker that he should buy a Windows license key instead of installing Arch Linux is worth squat.
It is literally impossible to do this. Revenue is the only thing that matters when measuring markets. If it doesn't affect the bottom line, it does not matter.
Passion does not sell OEM installs or license keys. Creators and hackers didn't buy MS products in appreciable numbers to begin with (compared to consumer and enterprise markets), so to the degree MS can be said to be losing them, MS does not care. If moms and humanities majors stopped buying laptops entirely and that $750M/Q dropped to the ~$1M/Q the power users drive, MS would probably pull out of the OEM license market.
This is a very different mindset about how businesses operate than you seem to share, which is fine, I'm explaining it not trying to convert you to it.
Anyhow, that's what Microsoft is abandoning. Familiarity. If young people do not learn MS Windows and MS Office during school or at home, the grip on the enterprise will vanish. We already see it in tech-oriented departments, where Macbooks reign supreme. I for one wouldn't like to use a Macbook, but I sincerely hope that MS grip on enterprises is broken sooner rather than later.
They never opened such panels, they don't read error messages, they just say "the printer doesn't work". If the "Add Printer" button doesn't work it's all over for them. The only thing you can do is make the "Add Printer" button as good as possible.
The least common denominator is so much lower than you think it is. It's why Jobs and Timmy Apple worked so hard to eliminate the concept of the "file system" and make everything built around the concept of the "app". These abstractions that seem intuitive to power users and technical people are nonsense to the much larger market of push-button users, and that's where all the money is outside B2B enterprise.
Apart from the issue of needing control panels mostly to solve problems, they're just not great.
Windows legendarily keeps multiple versions of them, while for ios/macos it's a game of finding which image is actually a clickable button or understanding that wi-fi hotspot is under sharing and not networks. The FA was about command line incantation to fix printers and of course at every new macos version there will be a set of new defaults to get back behaviors. I think that's just the way of life at this point.
Right now, we have machines that basically do what macs do (traditional "computing", a mouse, a keyboard a screen, office or image related work).
Then you have tablets (the whole Surface Pro line + 2 in 1 convertibles) and their touch options, including the impact on existing settings (you now have two or more sets of keyboards that might be in different mappings with different interlocked behaviors. Same for screens. and so much more)
Then thin clients, mirroring and remoting. For mac it's remote desktop or screen mirroring at most. Windoes gets a flurry of inbetweens.
Then all the hardware that only works on Windows and need some way to be managed relative to the system.
All in all, mac made the choice to only cater to the proverbial 20 of the 20/80 power law. Windows fundamentally can't just pile on the existing chaos, even if it means bringing new layers of chaos.
I don't even think a full rewrite in a new layer would be humanly possible at this point.
Not in the modern era, but I do love this one from Mac OS 1: https://i.imgur.com/XKfnGMl.png
Will the 10 person SME run such an setup? Unlikely. Can a 100 person SME that employs four MSCE run such a setup? Yes, but you only need two RHCE or LPIC-2 admins now (for vacation and somesuch). Those are rare, I know, but they will scale to 1000 persons, no problem.