While I agree with this view I am still mystified by the Rasmussen results. We have, at least in the past, trusted Raz to not cook the books. Of course we don't know what Rasmussen is using for a turnout model. It could be that he has made an adjustment to his R/D/I mix.
Or, maybe he has a pre-selected stable of LIKELY VOTERS who've been subjected to intense scrutiny and he just calls them up to see what they think this week versus last week.
Or, more likely, he pulls the results out of the air and takes a guess.