Please see my post above. Model II does not work on 2004 assumptions.
This is a radical re-engineering of polling by Gallup.
if you’re talking about the 49-46 model, then it assumes traditional turn-out which is the 2004 election as well...that’s all I meant...