SREFs have burned us many times. They can appear to be very stubborn which makes it seem that they have caught on to some trend ahead of the other models, but then it turns out they are just crap and were way off the whole time.
People also forget to take out the outliers (high and low) to produce the mean. There are some real garbage members in there (various ARW - Always Really Wet and NMM - Never Much Moisture runs) that always mess up the mean.
I would take the globals and smooth out the precip maps on them. They aren't going to be too good at nailing anything too specific in that regard. Snow max could end up anywhere from RIC to Mason/Dixon line.
Epic sick textbook snow pattern should be in place by 1 Jan... oops I mean 10 Jan... uh make that 15 Jan... never mind, try 18 Jan... sorry, I got it this time, 22 Jan. That's the window I am sure of it!