Change costs money? Yes it does; businesses buy the new Microsoft product because they changed the file types integration with Outlook.
Microsoft has to change things enough so that they can justify the "new thing" to sell -- and often that's a cosmetic change because 98% of their users aren't going to be using that new, deviation feature on a pivot table.
I could show a whole slew of changes Microsoft made that MUST HAVE had a lot of rationale associated with the change but makes no sense to me. Also a major complaint is why they changed the names of standard words in desktop publishing to useless and vague terms so you could never get a Help to find it for you. "Leading" is a term, "Paragraph Space" is a vague rearrangement of common speech. OK, rant over - the point is they started using terms for functions in their programs while ignoring real conventions. Then they made interface conventions based on some religion of UI while ignoring users.
I've spent many, many years annoyed at bad choices in UI by Microsoft. My fingers have had to travel MANY many miles because they couldn't have one alignment command with shortcuts and had to have an icon for every alignment -- they were ten years after everyone else to finally add a context menu so you didn't have to travel to center an object.
And Word still has major issues for anyone trying to do real page layout with it, or try and overlay high resolution graphics. It's a bloated text editor, no longer efficient and speedy at the first thing, and still hobbled by legacy to be good at graphics and page layout. But it's good at some email macros, and automation, as long as you don't have too many. Lot's of features, as long as you don't depend on them.
I just get really annoyed at bad designs -- it's not like I didn't KNOW how to use these product, or try and find better methods. It was the repetitive nature of using a damn baby hammer when I could see 50 ways they could have done it better. Why did someone get paid to make this lackluster app?
So now you are saying it's all driven by the touch screen. Well that's great, but I'm not using Windows on a touch screen and Surface is a tiny, tiny part of their market. Why not make a Surface interface for the Surface, and then a different one for the Desktop and marry the two when they've finally figured out the touchscreen and the desktop?
I'm not interested in researching Microsoft UI, because I've been too annoyed by their past efforts. Making icons big enough for a finger to touch them is NOT something I'd call a huge innovation, nor is it something that should have made Windows 8 such a bad hybrid.
I would really love to get into UI design -- actually "back into it" would be more accurate. But I did this stuff before they had degrees and certifications and NOW the UI is some sort of formal "it's done this way because WE KNOW these things." It's like the old "psychology for management" they used to teach in business school I suppose.
There are some things you can codify for "concepts" of a good interface, but it's really about a designer and a tool user crafting something that makes sense. There has to be a logic AND an aesthetic, it needs to let you know where you are and where you are going. It needs to be simple and elegant at the same time, and powerful should be one or two clicks away. It needs to be forgiving and let you know before you mess up.
It seems to me that UI designers are trying to create more and more UI, when really they should be like special effects in a movie; "There, but part of the story, and if possible, not even noticed." Strip away everything you can but no more from a UI and now you've got a tool.
Microsoft doesn't seem to have the mindset to to this; they look at the customer as something to market to and a touch screen as something to do LOTS of touching with. So I think whatever their UI is going to be, it's going to reflect the corporate culture, and then someone will "rationalize" in the new babble of UAX why it made kinesthetic logic to annoy everyone once again.