Although I can't tell you for sure how well the UKMET has done with Frances, I think that the model has had a right bias with Frances throughout the storm's entire life. Someone please correct me if I am wrong. So...the latest leftward shift is somewhat disturbing. This is the second or third run of the UKMET that has taken the storm more towards the left. Here's a good graphical representation of some of the models:
http://www.sfwmd.gov/org/omd/ops/weather/plots/storm_06.gif
You can see the UKMET model track, a track that will either cause Frances to make landfall in EC Florida or a very close brush. The tropical models, the two BAMM outputs, look alright track-wise but also seem to be a little slower. The BAMMs are better for storms in the deep tropics. The GFDL, well, is an outlier. Its track shows that the model may not have a very good handle on the high pressure system that will be anchored north of Frances.

So, there's still plenty of time to watch, but this storm has some hints of Floyd to me. If you have 7-day longwave spacing, then we should have another trough approaching the eastern US in another 7 days...next Saturday. How close Frances is along with her exact position will determine the effects on Florida, if any.