FWIW (from WND Daily) here is what is apparently one of the most accurate, albeit not scientific polls (although who knows-why should it not be termed scientific if it continues to be almost 100% spot on election after election?) which gives a remarkable result.
At this point in time it has President Obama being re-elected with one of the largest popular vote margin in history (approaching Harding's 26 point spread over Cox in 1920 and equalling Reagan's landslide over Mondale) and one of the largest Electoral College vote wins (Nixon/McGovern and Reagan/Mondale/Roosevelt/Landon were bigger landslides).
If the poll turns out to be accurate, then I have little doubt that the current result will have narrowed substitutional by election day. On the other hand, if Romney does very badly in the debates it is quite possible, given the current state of the traditional polls, that this result could happen.For those who decry this poll-well it will be right until proven wrong,which can't be said of many, if any, other polls.