I think we're arguing different things here. I fully agree that from a design perspective, things need to look distinctive (MacOS vs Windows vs Android etc). But implementation wise, the only actual difference is the images used and the colour of the gradient brushes.
The only shake-up UI-wise in the last 10 years was the introduction of smart phones, which meant the need to design for smaller screens, as well as work with gesture commands. Instead of adding to what we had, we spawned Bootstrap on the Web, Android has it's own custom widgets, as does IOS and so does Windows Mobile. This is the divided state of technology we live in. A smaller screen somehow communicated to people we need four new GUI frameworks.
This result stifles design, because if I design for IOS, I have to re-make everything again for android (including any custom controls, themes and so on). This has a cost, I can't support android UI controls if there doesn't exist the same on IOS, or other platforms. So people might choose to layer Xamarin on top, essentially throwing another framework into the mix. Why do we need so much abstraction?
Recently I've switched my desktop project from WPF to Eto.Forms, so it can also target Mac and Linux. But the result is I now have no mouse drag functionality, because that isn't implemented yet. These are real limitations I face everyday. What exactly is the problem I wonder? We are literally talking about squares with text/images on here. It is mind boggling.