The betting markets tend to give good predictions on the odds of a particular event occurring - much better than polling. Today was the first time that I saw the odds of Obama at below 50%.
It's not a surprise, really as the incumbent party's performance in the presidential election has been shown to be highly correlated (and dependant) on the economy's performance since 1970. In years when the economy is doing poorly, the incumbent party has always lost the presidency. When the economy has been doing great, the incumbent party has always won the presidency. When our economy has either been just coming out of a recession, or just going into a recession, it has been very close (with the incumbent party losing both times - 1976 and 2000).
I would guess that if the unemployment rate is above 8% next November, it doesn't matter whom the Republicans nominate - he/she will defeat Obama. If the economy improves dramatically and the unemployment rate is below 7%, it also doesn't matter, Obama will defeat anybody. The nomination process likely only will matter if the unemployment hovers somewhere in the seven-percent range next fall.