This has worked for them because the basic form factor of PCs hasn't changed in 30 years. They're still basically an evolution of a mainframe terminal but with the computer inside. Screen, keyboard, pointing device, etc. It's a very capable design but requires you either sit at a desk, or like laptops have a built-in desk that folds down and has an integrated keyboard and pointing device.
True mobile devices such as phones and tablets are very different beasts. They have highly restricted form factors and design parameters such as weight, battery life, screen size, finger-scale touch interface (as against more precise but inconvenient pen or mouse input). These restrictions demand a radical re-think of the UI that effectively closes off a lot of options. Trying to cram in those options anyway, no matter how inappropriate they are to the form factor or how much they compromise the user experience, shows an appalling lack of discipline.
Microsoft looked at the iPhone and iPad and completely failed to understand that it was the clarity and focus of the UI design that made them successful. That it was precisely the jettisoning of desktop UI metaphors and user interface cruft that made them so appealing.
So what Microsoft did was develop exactly such a no-compromises built-for-mobile UI in Metro, and then bolted it on to the side of a crufty old desktop UI, with an array of laptoppy peripherals. They also decided that while the system launches into he new UI, you actually have to figure out how to trigger the invisible gesture to get to the old UI to do anything useful like run Office or change basic system settings. I honestly don't know how they could have botched the whole thing more completely.
So yes, you're right that Microsoft have stuck to form an maintained backwards compatibility at all costs. The problem is that this time that cost is a viable place in the mobile market. The thing is, they don't need to do this. Desktops and laptops are having their market share adjusted downwards to make space for true mobile devices, but they're not going away. Anyone who wants to run traditional Office or legacy DOS applications can still buy a device to do so. There's just no need to the new generation of mobile devices to have to cater to a market that's already catered for. Suppose Microsoft had launched Surface last year with no desktop, Metro only and with a native Metro version of Office and Outlook. I think the situation now would be very different. Maybe that would have been impossible given the time constraints, but that's what needs to happen.
True mobile devices such as phones and tablets are very different beasts. They have highly restricted form factors and design parameters such as weight, battery life, screen size, finger-scale touch interface (as against more precise but inconvenient pen or mouse input). These restrictions demand a radical re-think of the UI that effectively closes off a lot of options. Trying to cram in those options anyway, no matter how inappropriate they are to the form factor or how much they compromise the user experience, shows an appalling lack of discipline.
Microsoft looked at the iPhone and iPad and completely failed to understand that it was the clarity and focus of the UI design that made them successful. That it was precisely the jettisoning of desktop UI metaphors and user interface cruft that made them so appealing.
So what Microsoft did was develop exactly such a no-compromises built-for-mobile UI in Metro, and then bolted it on to the side of a crufty old desktop UI, with an array of laptoppy peripherals. They also decided that while the system launches into he new UI, you actually have to figure out how to trigger the invisible gesture to get to the old UI to do anything useful like run Office or change basic system settings. I honestly don't know how they could have botched the whole thing more completely.
So yes, you're right that Microsoft have stuck to form an maintained backwards compatibility at all costs. The problem is that this time that cost is a viable place in the mobile market. The thing is, they don't need to do this. Desktops and laptops are having their market share adjusted downwards to make space for true mobile devices, but they're not going away. Anyone who wants to run traditional Office or legacy DOS applications can still buy a device to do so. There's just no need to the new generation of mobile devices to have to cater to a market that's already catered for. Suppose Microsoft had launched Surface last year with no desktop, Metro only and with a native Metro version of Office and Outlook. I think the situation now would be very different. Maybe that would have been impossible given the time constraints, but that's what needs to happen.