This is really amazing stuff. The match sample sizes are enormous in many cases, giving far more credibility than we could ever get from MTGO challenges alone.
Kevin and I didn't do a podcast preview show this year, but we did spend an evening documenting our predictions for the event (which was recorded in a google document). I don't want to reveal Kevin's, but my final prediction was:
3% Big Blue
3% Dark Ritual Combo
4% Blue Control
I'm in striking distance every year, but this may have been my best estimation ever.
Re: the metagame. Not only do none of the major decks have dominant metagame presences, but none have dominant win percentages either. This is a very balanced Vintage metagame. I am enjoying it greatly.