I've worked as a software developer at large companies. Usually, "Good enough" is all you can shoot for. You aren't given the time nor incentive to make something truly great. Not that any one person really could, even if they had the motivation. Do you know how many lines of code are in Windows 10? 50 million. That’s 50 million lines of code written by hundreds or thousands of different people, some of whom are most certainly dead and didn’t comment their code properly, over the span of decades, since I guarantee you Microsoft wouldn’t start fresh with their flagship product. There’s still code in there from Windows 95 if for no other reason than to support some legacy compatibility modes. No one person fully understands it. To be honest, I consider it a miracle that Windows works as well as it does today. I can count the number of BSODs on my custom built gaming PC on one hand. That is an astounding achievement.
To fix this problem, you said you needed experts. These experts will never exist, because the amount of time you need to gain enough experience to solve this problem is greater than the length of a career as an OS developer, designer, tester, or manager. Ramp up time would be a nightmare for any team. You’d have to pass knowledge along to others, and even with the best documentation in the world, ramp up time means that it’s simply not worth it to solve this problem for any software company, even Microsoft.
I also want to say that a perfect system like you want is impossible without knowing EXACTLY what you need right from day 1, which has never happened in software development EVER. Have you ever played Factorio? You always hit a point in making your factory where you think "If I had known I needed this at the start, I would have done this section totally differently. I can either tear down EVERYTHING or accept some inefficiency." Humans, being fundamentally lazy, pick the latter unless the problem is truly too big. Requirements for software change. Constantly. You get a new customer who wants the product to do one more thing, one more feature. You didn't plan for it, and it causes some small slowdowns, but hell, let's do it. It would be cool. Or you get a security issue in that cool new thing you designed. Or the head designer who didn't take good notes dies in a car crash, or one of a million other things goes wrong. Or you're asked to write a fix on a section of the product you don't understand because the usual guy is sick and can't come in, and no, it can't wait, so you write a hack, it fixes the issue, and nobody looks at that section of code again.
Tl;Dr: The nature of software development is such that "the perfect GUI" that you want will never come into existence. Especially not from Microsoft.