hi dig,
i think the republicans think that if the pollsters find more democrats when sampling the population, this means the sample is skewed.
the simpler explanation. there are more democrats. the poll did what it was meant to do. take a random sample using a particular methodology and measure the results.
the only things to iron out are to do with preferences of medium and suchlike. if you only robocall landlines, then you create a result in line with political preferences to do with landline use and patience with robocalls.
sort that out and you have a result.
but really, this is all about republican worries about confirmation bias. once obama is seen to be the favourite, people will prefer him and their coffers will run dry. nothing to do with the polls.