It's really hard to say overall. I prefer the Euro for the most part on the larger synoptic patterns. But more than once in the spring the GFS has big cold fronts correct, even 10 or 12 days out. While the timing would change by 6, 12 or 24 hours it handled some significant patterns rather well.
Each system is different and what has me intrigued on this in particular is the nature in which the GFS really isn't wavering much. It's been generally consistent for 10 or 12 runs in a row now with timing bobbling a bit here and there and on some runs not as much cold as other runs but still generally much colder than average east of the Rockies. It may be hour 96 or so before we start to the see the model gap narrow.