I'm pretty sure Rasmussen uses a statically weighted model. e.g., the D/R/I percentages are always fixed... see my earlier post... he would/could not have predicted 1994 or 1984 blowouts, and he would've most likely messed up with Perot in 1992 as well.
My understanding is that he does: D 39%, R 35%, and I 26%. This is second-hand, but I have seen more than one person say it.
My own attitude is that any poll that weights the samples by party affiliation is rendered unscientific and invalid for that very reason. It destroys the randomness of the sample, especially of likely voters. By contrast, Gallup doesn't weight their polls. They don't make the mistake of assuming that just because, say, 39% of registered voters are registered as Democrats, 39% of likely voters will therefore be Democrats. Sometimes you can lead a donkey (or elephant) to water, but you can't make him drink.