I highly doubt that non-registered voters were allowed to vote. Also, it isn't possible to have more people vote than there are registered to vote, unless you let someone vote multiple times. It would be interesting to see real proof, not a poll, to substantiate this claim.
One of the primary purposes of Polling is to put a check and balance detector on the possibility of massive voter fraud.
Just because the results of the election from any given pollster does not match the given election outcome, does not automatically mean that the particular Pollster is wrong, it COULD mean that this is place were voter fraud was a significant factor.
Let me say it again, if a Pollster's predictions don't match the outcome, he isn't always Wrong!
It could have been Election Cheating!
One way to test this theory is to try to match Exit Polling and Outreach Post Election Vote Declaration Polling to the listed election outcome.
We all know that putting a Republican Support Bumper Sticker on your car is a great way to get your doors keyed, tires slashed, windows scratched or shattered, or if an OWS guy is nearby, have your car pooped on...
In a Similar Fashion, Republicans have learned the hard way, DON'T talk to exit polls.
Yes, Voter Intimidation is a REAL Problem in America.
So that leaves Annonymous Post Election Voter Declaration Polls as the only real check without significant bias.
There hasn't been time enough to conduct such polls, and there needs to be a bit of time between the election and the Post-Poll to get people willing to Talk.
It will take some Time for the picture of the level of voter fraud impact on the 2012 election to become clear, but already we see it was a major, if not determining factor.
-