I remember a storm not so long ago that the GFS showed as a blizzard in NC, and it was on an island. I think we know how that turned out. Any time a model is alone in showing an extreme solution, it's almost certainly not going to be right. Unless that solution is an extreme drought - then it's always right.
I realize I'm stating the obvious, but I'm doing it as much for myself as anyone else, and I sure hope I'm wrong.