Another interesting thing is that interaction with land can always affect the track. If I recall from 2005, the forecast on Katrina was not very confident until it reemerged off the coast of Florida. Also, if you look on the 400 - 850 MB Layer Mean Wind Analysis, it seems that if Dean were to meet up with the system over the Bahamas, or it deepened, Dean could very well be pulled north. So basically it seems that the soon Dean were to seriously interact with the system, the farther north he would travel. Also, historically speaking, it seems that the National Hurricane Center forecasts for days 3 - 5 tend to be too far to the left (i.e. west). Examples of this include Ivan in 2004 (originally forecast to hit Biloxi, MS then hit Pensacola, FL), Rita in 2005 (originally forecast to hit Port O'Connor, TX then hit at the LA-TX border), and Ernesto in 2006 (originally forecast to hit the Florida Panhandle then hit the Miami Area). The northern Gulf for sure cannot be said to be "in the clear" at this point.

Jason
http://www.freewebs.com/jays_weather_center