I don't have time to do this now, but we need to calculate hurricane effect on 2005 figures to get a better idea of how far off IEA really was. That said, we also need to remember that this is a real effect and could easily repeat in 2006 and into the future, at least to some degree, given the hotter Gulf temperatures these days.
A rough approximation of the impact of the two hurricanes in terms of decreasing global production is about 250,000 BOPD average for all of 2005.  This assumes about 750,000 BOPD shut in (on average) since September.  This is probably an overestimate of pure crude, but I am not sure how much condensate or NGL's come from the Gulf, and a lot of natural gas is still shut in.  

This rough forecast also does not take into account any additional production that would have been brought on if the hurricanes had not happened (e.g. starting up Thunderhorse).

As for your "rough approximation", Bubba, I'll buy that and it seems to be reflected in IEA's Figure 1 in my post. It seemed too hard (in other words, I was too lazy) to figure out the actual effect on world non-OPEC production resulting from the hurricanes.

However, I think this relatively small "oil shock" hardly changes my analysis of the IEA forecasts and revisions over the last 5 years. And, as peakearl implies, perhaps the IEA ought to factor in the risk of such Gulf of Mexico production disruptions each year in making their forecasts.
Thanks. Seems pretty reasonable to me.