This is a good question. I think the answer is two-fold. First, the longer ranges models that most of us can access without having to pay (GFS, GEFS, GEPS) very frequently show below to much below average temps at range during most winters. This is a pretty well documented bias. Therefore, it's wise to incorporate that bias into one's expectations. The EPS isn't as bad in this regard, but most of us don't have access to it unless someone posts it for us.
Secondly, for whatever reason, the majority of winters of late just want to be warm. Maybe there's some cynicism mixed in, but the fact of the matter is, it's easier to be warm of late than cold. Given that, a model showing cold, while nice to enjoy, just given persistence (and bias as noted above) seems less likely to be right than one that is showing warm. That doesn't mean it always works out that way, but when you experience something like this over and over and over, you sort of develop the model that cried wolf mindset.
One day, hopefully, this will change. But until it does, it's probably wise to be skeptical of longer range cold patterns showing up.
