First off, a huge thank you to everyone for all the kind compliments - coming from denizens of this board, this is high praise indeed.
Second, to answer the folks asking about data from earlier years, the link I posted in my first comment contains data for 2008, 2004, and 2000. While we only compiled 2008 ourselves, the older data is from a trusted source (Polidata by way of the Cook Political Report) and was used in the Almanac of American Politics. Please note that the 2000 data was indeed recalibrated for the post-2000 census redistricting, and the 2004 data for Georgia and Texas reflects the mid-decade redistricting those two states did.
To answer the question about early votes and split precincts, I don't think that those states (eg, Georgia) ever release proper data. We've made some educated guesses about what those splits are or ought to be, and we will probably have a more detailed post forthcoming about that. While it would be possible for results to shift somewhat, we think it's unlikely that there would be any major surprises.
Awesome. Perhaps you can inform Dave Leip and see if he would permit the 2004 data and maps to be added to this site?
