I agree to a point, but actually Microsoft wouldn't have to be heavily involved really. With ODF, that push didn't come from Microsoft and eventually landed with standards bodies. Microsoft fought it all the way, and even though they muddied the water with OOXML, they had to stop implementing binary, opaque and hard to manipulate file formats.

They were then forced (by market demand) to implement ODF. I can see something similar happening in terms of document automation: only maybe this time it might be quicker given there is experience around the DOM in web browsers already, so a model is already usable.