re: And ON2's codec is more likely to be accepted by everyone than Ogg because...?
Because google will get grown-ups to work on it, likely.
If you've ever wasted time on the ogg/theora codebase, you'd see it was a messy, crabby poorly-documented and ass-backwards implementation wrapped around a meh-quality at best (because it's old, was good in its day) coder written by someone else.
It's not nice to develop against xiph's libraries for audio or video- I can't blame a lot of people for preferring libavcodec/ffmpeg for open source stuff and sticking fingers in ears over licensing issues.
If Google were to take a later version of the On2 codecs than xiph have, and offer sensible specimen implementations, people might just go for it. A decent codec and a decent implementation could be quite tempting.
Whatever your stance on Google is, they certainly Get Stuff Done(tm), so they could make this one fly if they wanted to- a bazillion dollars > a few sulky, incoherent volunteers (having interacted with some of xiph.org's more senior developers, when offering to fin their development of open source, I was not impressed).