Where do you find MS's greatest successes? MSDOS was built on top of QDOS, purchased from another company. Windows followed the MAC and Lisa OS (And the Xerox Star). IE really followed Netscape browser. The X-Box followed generations of game consoles.
I'm not making a statement about "good vs. bad" relative to this approach, just that following and improving is what they have done best. That's where MS has succeeded to the greatest degree in the past.
The have the money, person-power and marketing muscle to look very closely at what others are doing, see what customers do with it and then build a product that will work well for the masses. They haven't always been successful at this approach (phones, music players), but if they keep at it, they will build a broadly used product.
Given that, I can see them dabbling in ARM product OSs for a while and then diving in full force once the market settles into real-world use for the products. The past few years as well as what happens with ARM OSs in the next few years is really not relevant to MSs ultimate success in that market.
Medical has the same barriers that defense and industrial have - All of these demand certification of reliability. In the past this has not been a Windows strong suit. Maybe it is more achievable now, given the emphasis they have made on security and reliability over the last couple of years, but it is yet to be proven.
Given that cost is a major factor in most consumer purchases and ignoring the dedicated high-tech savvy minority (the early adopters, no price too high tech types) the way for MS to make inroads is to provide low cost ARM based Windows products. I would suggest a streamlined, no-frills OS with basic word and internet capacity. Considering how many devices are going to be in the medical category, MS would do well to target ARM based medical devices. With the medical market, cost is not the main factor; the main concerns are reliability, robustness, and intuitive user interface. MS could do well to be in that market.
You can probably argue this on the desktop, but I'm found WinCE no better than using an embedded linux. The online support for embedded linux is just as good as WinCE and I've found porting linux easier.
So in the embedded world, which is closer to the tablet world than the PC work, Windows has no advantage.
It will not be long for people to realize that there are better (and free) OS than Windows OS. It will happen with the end of era of PCs. Google is already tapping market for mobile OS and web-based OS, which might change the way we will work on mobile platform.
Microsoft may be slow in bringing out new packs but their software packages are very efficient ,user friendly,reliable and they have a good on line support.So MS will have a broad range of users for ever.
Windows can go into many devices if Microsoft comes out into market with a new version of operating system for different class of devices such tablet and PC. But I feel they are slow and non-innovative. Never the less I think it is time for Microsoft to show some new colours just like what google did with chrome and android.
I agree with Rick's opinion that Windows needs to support ARM in the server side. The Windows licensing fees are not going to fly vs. Android, and they won't take any seats away from iOS :). Why not defend the home turf where people are still paying for an OS and you are still a market leader not a 2nd tier me too OS.
"Maybe MS could create a streamlined version of Windows and Office for the tablet, one that ran fast and wasn't such a memory and power hog, ditching many
You just described Open Office...
Here lies Microsoft's problem, actually: they put together a well-integrated platform (Windows + Explorer + Office), made sure that it's not modular, for technological and legal reasons (remember how they argued during the anti-trust trial that IE cannot be separated from Windows?), and now it bites them on the mobile platforms.