Question- Does Rasmussen apply their own findings of Republican registration advantage of +4%? If they do, their results would seem to pretty bleh for Romney. Of course, there are other elements to a secret sauce, relating to turnout by age, gender, religion, ethnicity, etc. separate from party ID. Actually, while I'm a fan of Rasmussen, I don't see them listing their actual internals. Am I missing something? How about Gallup's party ID?
Seems to me that in addition to asking about gender, age, and 50 different policy questions, every poll might want to inquire as to how their respondents actually voted in 08, as well as party ID, and adjust for aging factor accordingly.