I think it’s more that long range ensembles typically show a pattern washed out in space and time, where there’s typically more variability, especially in winter, due to storm systems that aren’t seen in long range ensemble means. As you get closer, the ensembles resolve that variability. For example, next week still looks BN overall, but it was looking like wall-to-wall cold a couple days ago. Now you’re seeing the ensembles pick up on a brief warming around next Wednesday due to SW flow ahead of the next cold front.
That said, the gfs and GEFS definitely have a cold bias in the long range, and there’s also a tendency to overdo HL blocking.
I’m buying the cold look more right now because the Eps is very bullish on cross polar flow.