Not a Google engineer, but I can offer a partial explanation.
As you probably know, up until the release of IE9, supporting apps across the major browsers (Safari/Chrome, Firefox, IE) was a massive PITA. Not just obscure features, either; even IE8 had some crippling misunderstandings of the CSS 2.0 standard (floats). Today desktop cross-browser problems are consistent subpixel antialiasing and less-critical CSS features like border-radius support, so it's pretty easy for an app to [mostly] work in every modern browser without testing. Plus computers are fast enough that you don't [hardly] notice IE10's stupidly slow JS engine.
On mobile, that's not the case. Web-based Google Maps taxes my iPhone 4S Safari to the limit. It's so jerky it's hardly usable. V8 (Chrome's JS engine) is faster, so Android users might have more luck, but you're pushing your phone's hardware regardless. I haven't had the misfortune of owning a Windows Phone, but knowing Microsoft's previous disregard for standards, there may be some significantly broken features of IE10 mobile (though it supposedly supports HTML5 to the same extent as its big brother). At the very least, IE10 sees fairly significant lag with a few tabs on a recent desktop computer; I imagine JS performance on a phone is horribly unusable for an app like Google Maps. Optimizing JS for IE10 mobile could be a non-trivial task with diminishing returns—it might not even be possible to get it to run acceptably.
So there's a chance that capturing the 2% of the mobile market on Windows Phone isn't cost-effective for Google. It's probably political though—demos of IE Mobile have looked decent. [1]
Hopefully a real Google engineer can provide a more definitive answer though.
As you probably know, up until the release of IE9, supporting apps across the major browsers (Safari/Chrome, Firefox, IE) was a massive PITA. Not just obscure features, either; even IE8 had some crippling misunderstandings of the CSS 2.0 standard (floats). Today desktop cross-browser problems are consistent subpixel antialiasing and less-critical CSS features like border-radius support, so it's pretty easy for an app to [mostly] work in every modern browser without testing. Plus computers are fast enough that you don't [hardly] notice IE10's stupidly slow JS engine.
On mobile, that's not the case. Web-based Google Maps taxes my iPhone 4S Safari to the limit. It's so jerky it's hardly usable. V8 (Chrome's JS engine) is faster, so Android users might have more luck, but you're pushing your phone's hardware regardless. I haven't had the misfortune of owning a Windows Phone, but knowing Microsoft's previous disregard for standards, there may be some significantly broken features of IE10 mobile (though it supposedly supports HTML5 to the same extent as its big brother). At the very least, IE10 sees fairly significant lag with a few tabs on a recent desktop computer; I imagine JS performance on a phone is horribly unusable for an app like Google Maps. Optimizing JS for IE10 mobile could be a non-trivial task with diminishing returns—it might not even be possible to get it to run acceptably.
So there's a chance that capturing the 2% of the mobile market on Windows Phone isn't cost-effective for Google. It's probably political though—demos of IE Mobile have looked decent. [1]
Hopefully a real Google engineer can provide a more definitive answer though.
[1] http://thenextweb.com/microsoft/2011/02/18/watch-internet-ex...