Certainly no argument with that.
But I think these past few years there has been way too much emphasis placed on face-value output of the ensembles without taking a step back to fully assess. I mean how many times over the past few years has the medium-range pattern looked great only for the complete opposite to happen at verification time?
IMO, when you have these Nina conditions where the global pattern really is quite chaotic with so many moving pieces and so many ridge/trough patterns globally, sure the ensembles will probably pick-up what the pattern will average out too, however, it's going to fail at picking out the short-term variations within those patterns. I think this is why we've seen medium range guidance look great for snow events, but once that time comes its a crappy event b/c the way the pattern is evolving around the storm development its complete crap.