The iPad (and HTML5) did kill Silverlight and Flash (to some extent) as browser plugins, but not as platforms in general. Flash seems to be hanging in there outside of the browser plugin space, and Silverlight died because Microsoft thinks it's more fun to respond to market challenges by sweeping everything off the table and into the dustbin than it is to just pivot and keep moving forward.

Here's an alternative strategy that I strongly suspect would have been less expensive for Microsoft in terms of both development cost and social capital: Keep Silverlight as the mobile development API, and carry it forward as the platform for developing on both WP8 and Windows tablets. Pay Xamarin to incorporate support into their dev tools. Deal with performance considerations by expediting the AOT compiler for .NET, which clearly ended up needing to happen anyway. Deal with cross-language goals by . . . wait, you're still on .NET, cross-language is one of its core competencies. No need to sink resources into developing yet another iteration of COM, so that's a huge time saver. And as an free bonus, take advantage of the fact that Silverlight's already got an OS X implementation to let your developers use the technology to ship cross-platform desktop apps, similar to what Adobe's doing with Air.