I've never been a big fan of the 18z/1pm off model run and even though the GFS did make it colder and a bit more snow on the northern fringes we need to see a bit more on the run tonight.
Mike DeFino's blog pretty much spells out the errors on this run. The real precip (radar) is more north than the model run and real precip will be a predictor more than any model guidance.
If you take the GFS and by the way that is my model of choice...you get a decent snow for Philly south and east plain and simple. There still may be a sharp cutoff as very cold air at 5000 feet comes rushing down. That could also help on the fluff factor for the fringe locations like Philly.
The one thing we have never done on this site from the pro met side is herky jerky forecasting due to a model run or one model...just not how it works in large weather systems and model trending. For now I keep everything the same as the NWS has posted. The wild card will be the radar for the overnight and the model run trending at 0z/7pm tonight.
My 2nd call is about the same as the one earlier so here goes...