Santorum is going to get a share of that Evangelical vote, which will probably be around 60% of the 08 Huckabee vote. When I started working on the final simulation last night, I originally was going to follow a 60-40 split for Newt, but as soon as I applied it to the major counties statewide, Newt was getting too much, like over 45 and closer to 50%. He was coming within 50 votes of winning Charleston and 125 of Beaufort. If that happened, the Romney people would not have even been in SC today, they would have packed up, conceeded, and moved to FL.
Also, Women in SC make up about 48-49% of the primary. More Women voted for Huckabee and McCain last time than Men, getting 33% and 34% respectively. To give Newt a majority of the Evangelical vote would mean giving him a near plurality of the Female vote in SC, something which is very unlikely to occur at this current time.
Even as my model stands now, I have my doubts about Lexington County. This is considered a favorable area to Romney that has Gingrich barely winning, but still, a win here is a big thing.
The work you did on that was astounding, but there is a new dynamic this time that I don’t think your modeling accounts for (though I may be wrong).
That is another 10-12% of raw vote total. (445K last time, maybe 500 K this time).
This will come from intensity - folks who were not there last time, and I think they will be Paul supporters or recent Newt converts.