I was looking back on Nemo from 2013 right before the storm got going. The GFS was showing the storm well SE of where it actually was on radar. Point is, when we're getting this close, accuracy on the major models isn't really as important.
If I do remember correctly, I was watching Joe fury the night before talking about some the NAM output for snowfall accumulations, and Joe was saying he's not sure he believes what it was projecting. We all know where we ended up in Connecticut. So the Nam snowfall projections wound up coming true.