I like the Idea, but must say I would love to see how its going to work out in the end. Reasons I reserve judgement:
I dont know how they going to handle 3rd party apps as you wont be able to run Photoshop or Crysis on a arm based Windows 8. There is no way in hell an application compiled for x86 will work on ARM. Microsoft themselves is a bit closed lip about this part. You will not even be able to emulate X86 on ARM so only way to do this is to emulate ARM on x86, and that will come at a huge performance dip.
The second part of the equation is what about the people that dont like the new interface, people love to customize their PC's especially gamers. I know you will be able to switch to the Windows 7 look vs Metro, I hope there is also hooks into the metro API to allow for custom theming. Personally I hate the blocky 8 bit look of Windows Phone 7 and yes I have a WP and Metro look is the only reason I actually hate the phone. Also dont get me started on the words that run off the page. For a dyslexic its pure hell.
Indeed, the points you raised are exactly the type of things I'd like more clarification on instead of those redundant and bland opinion based articles.
I think the crux of the whole thing lies in the way Windows 8 might or might not support "legacy" x86 apps. I'm not so concerned about the way Windows mobile UI currently looks and works, I don't have a Windows phone so you'll obviously know more about that.
To me that's something they can always change or adapt as time goes by (or hopefully, allow others to change through some decent API's..../EndDayDream).
I really do hope they manage to pull off legacy x86 support within acceptable power usage and performance parameters, not because I support MS beyond reason, but because it's something I could truly use.
What's woefully missing from a "hype" point of view from all the Windows 8 articles I've read is the observation that IF they pull this off somehow, their competitors are kind of screwed. You can always change, adapt, update or extend the functionality of the "mobile" aspect of their OS, but you can't just go from a truly mobile OS like Android or iOS to something that rivals Windows 7 functionality and the massive established application base behind it.
Even if they fail, there's probably still an upside to this... Apparently they've had to put a lot of effort into optimizing the Windows 7 code base (which most people will agree with, is already quite brilliant in terms of performance and stability).