It is probably too large for them to be able to determine if it can be efficiently implemented in DirectX with hardware acceleration. In order to know that, the

It is probably too large for them to be able to determine if it can be efficiently implemented in DirectX with hardware acceleration. In order to know that, they pretty much have to implement the whole thing and see. And, if they find things that are poor performing (i.e. its primitives do not line up with DirectX's primitives), they're going to push for changes to the spec to make it efficiently implementable. But, those change proposals are likely to fought by the other browser makers. Also, I believe Adobe is likely to ask for changes to the <canvas> specification too. Plus, <canvas> is not that important if you have a high performance SVG implementation. Finally, <canvas> creates a situation where animated advertising cannot be easily distinguished from other content and cannot be separately blocked from other, useful, Javascript on the page. So, I think there's more uncertainty regarding the <canvas> specification than people might expect. It makes perfect sense for everything to become more clear before committing to support it, considering Microsoft's very generous support and backwards-compatibility guarantees to its customers.