Instead, it's the projection of the general election that is interesting. It seems to suggest a wave for the Democrats bigger than 2006 or 2008.
The facts: Poll taken over 3 days, 12/4-12/6, a short time for a college poll. 2,107 registered SC voters, MoE 2.1.
Again, I'm not even looking at the actual results much, though they do show Newt Gingrich solidly ahead in the three-way matchup against Mitt Romney and Ron Paul, up 48-30-12 among 635 primary LVs (MoE 3.9). I care about the general election sample.
According to Marist's "Nature of the Sample" table, Democrats make up 33% of the sample of RVs, Republicans 30%, and Independents 36%. That sounds reasonable at first, but this is South Carolina. At the peak of the wave for the Democrats, the 2008 election, CNN's exit polling shows a breakdown of 38% Democrats, 41% Republicans, and 20% independents. Independents appear to be vastly oversampled, and taking more from Republicans than from Democrats, in the Marist poll.
Democrats do traditionally have party identification advantages in America, but even in a great year for Democrats, they failed to have one in the 2008 elections in South Carolina. For Marist College to project such a swing in 2012, they must be predicting an even bigger wave in 2012, than the Democrats had in 2008.
Or we could just dismiss the poll as premature and likely wrong. That's where I stand.