But apps that run on iOS don't run on Android or vice versa. There are 7 billion people in this world. Knowing that I can't possibly reach all 7 billion people is nothing that I'm too concerned about. You've got to pick your niche when it comes to software, and if the niche is defined by not only the domain only but also the platform, then so be it.
Which is why the HTML5 announcements are hot air, why make a song and dance on something most web developers are likely to ditch for a browser because of reach. Even the interns that made the app at build will either get work at Microsoft, or ditch their HTML 5 efforts, for a website that can be consumed on Windows, IOS and Android.
I don't know why doubling their resources is necessary. They obviously devoted considerable resources to WinRT and tooling for Metro-style apps while neglecting WPF, but Silverlight recently received a refresh with its new out-of-browser capabilities (i.e. child windows, etc.). It looks like Silverlight is intended to replace WPF. Everybody is asking "Is Silverlight dead?" when the question to be asked should be "Is WPF dead?"
This just isn't a framework thing, they need to continue to make the desktop exciting, and revolutionary, for the bulk of users that don't want the tablet UI forced on them. Windows is quite a nice place (especially Windows 7). I'm not quite sure what you have smoking away in your pipe saying that Silverlight would replace WPF when Silverlight is in fact WPF/E (everywhere), its a completely absurd assertion.
The software world has moved on with Apple and Google gaining market share, and Microsoft still trying to use the same amount of resources to develop desktop, server, and the mobile tablets and phones. They are starting to look stretched resource-wise and unable to adapt or keep up the pace.