Well, you're both right in different ways. The ensemble suite is useful for gauging the stability of the evolution in the short- to medium-term, ie perhaps out to 4 or 5 days. It is beyond that the ensembles can't detect something unexpected that happens in the atmosphere: clearly things do change otherwise the forecast charts would be rock solid for a much longer period of time. We have all seen the ensemble suite flip unexpectedly; it's happened in the last couple of days, as Brian commented.
As Rumsfeld said, in a different context, there are known unknowns but there are also the unknown unknowns. It's the latter that brings the uncertainty.
Originally Posted by: Gandalf The White