[BillC]

David,

I think your bet is a bad one. Pascal wouldn't even make such a bet. The proof is correct. It is perhaps the most scrutinized big theorem around. This only comparable modern mega-result is the classification of finite simple groups (which is much much longer...).
Plus, you have to worry about bankroll concerns if you get much action. See my July magazine article on factoring in variance for longshot (high variance) bets.