At times, when the NHC or models have come under fire, I stated that the error in the 5 day forecasts are now what the 3 day forecasts were 20 years ago. That may have been a conservative estimate.
According to this article, the average 3 day error by the NHC has decreased 50% since 1998 alone. They do talk a little about the new Md. model also.
I will not blast the NHC for Charley, they had warnings up for where it hit. The media always focuses on the worst-case scenario.
In this last storm, I will not blast them for their errors due to the size/strength of the storm, and they eventually did get it right. Because it went farther west than everyone thought, that doesn't make the western-most models good at forecasting hurricanes, any more than the eastern most models bad, either. In fact, they stayed too far to the left until it was nearing the coast. The interesting thing is, if we were our parents age, (for those of us that are out of school), there were no models to criticize, therefore, no discussion. I bet they wish they had them.

http://www.newsmax.com/archives/articles/2004/9/16/84808.shtml