It's no secret that Intel has found little success in the mobile market, considering they are typically behind in performance/price/battery life ratio compared to ARM chips, and their Atom chips are also typically a generation behind in both CPU and GPU performance, too. Therefore, OEMs have no reason to pick Intel over any of the much cheaper ARM chips out there.
So what's Intel's solution to that?
They want to push Android to Windows machines, because then they can market themselves as the "only company that can make this happen", since Android is architecture agnostic (meaning Intel's chips work with it, too), and the full Windows only works on x86. Yes, there's Windows RT, too, which OEMs could use together with Android with ARM chips, but even fewer people want Windows RT than the full Windows 8.
But why isn't this a good solution?
More is better, right? Not always. In fact, Windows 8 alone shows why more is worse. They've stuck together two very different operating systems, or at least 2 very different user interfaces, which seem to confuse the hell out of normal people, especially those used to the old desktop interface, because Microsoft keeps designing Windows in such a way that they continue to force you to use the "Metro" mode. And now Intel wants to add a 3rd OS/UI paradigm to the mix? Yes, I'm sure what won't confuse people at all.
Dual-boot – higher price
One of the problem with such an Android/Windows system is that it drives up the cost, too, and ends up costing more than it would to buy an Android-only equivalent, in part because Microsoft tends to ask for at least $100 for a normal Windows 8 license, or even $200 for a Windows Pro license, so that can make quite a bit difference on the final price.
Intel will also make it so Android runs inside Windows in a virtual machine on these devices, which either forces the OEM to add higher specs to the device than it would've otherwise, just to keep up with the requirement for virtualizing Android inside Windows, or Android won't run very well inside Windows – especially if the devices has a high resolution, and Intel isn't cheating by making the Android version run at a much lower resolution.
Besides the Windows license, and the higher system requirements to run this set-up, there's also the issue of support. Companies don't even want to support Linux on their Windows machines, even when they are selling them with FreeDOS on them, in some countries, because they know people don't want to pay for the Windows license there. They could at least put a real Linux distro on them, but they do not, because they don't want to support it. Let's also not forget how poor Android OEMs are with Android support, too, for most of their devices.
Even if Intel is abstracting all the necessary components to run Android on all machines the same way, it will still need to be supported by them, to make sure it's not crashing on certain devices and so on, which at the very least increases the cost to Intel, but it will most likely translate to higher prices for those Intel CPUs inside these devices.
Dual-booting is bad for both Google and Microsoft
Microsoft still owns the majority of PC market share, but its share in the whole "computing" market is much smaller when you add Android and iOS. They would obviously want users to only use Windows and nothing else. They don't want people to learn to use Android. Putting Android on Windows machines means more people will get used to Android – worse yet, they will get used to Android even on desktop PCs. Somehow, I find it unlikely that Microsoft likes this idea at all.
Google owns mobile right now, at least in terms of market share. Android is on the majority of smartphones and tablets, and I'm sure they have high hopes for other computing devices, too, including PCs. So why would they want people to stick with Windows, when Android has so much potential on its own? Google didn't even want to make Chrome OS a "secondary OS" on Windows machines. They certainly don't want Android to be a "secondary OS" to Windows – in any market. This would only benefit Intel.
People who want Windows are going to get a Windows computer. People who want Android, are going to get an Android computer. It doesn't have to be anymore complicated than that. I strongly believe Android will become such a mature OS in the next few years and with such a large and advanced app ecosystem (significantly bigger than any other app ecosystem out there), that there won't even be discussions about which is a "mobile OS" and which is a "full OS". Some will stick with Windows or Mac OS and iOS, and most will be using Android, for all of their computing needs.