The 12Z models are all in and the clear message is that areas NW of 95 are likely to get less precip/snow than people along 95, who are likely to get less snow than folks towards the coast, partly due to less precip and partly due to where the heavier snowbands and the light subsidence zones set up (which is NW of the heavier bands). I think Steve D.'s map above captures a lot of this well, as does his discussion in the video.
Based largely on the 12Z models, one could imagine the 2" line being from about Reading to Allentown to Hopatcong to Newburgh and the 4" line going right up Route 1 from Philly to New Brunswick to Newark to Yonkers and the 6" line going from Hammonton to Long Branch and across to Hempstead on LI and 2-4"/4-6" being between those lines and <2" NW of the first line - and 6" being the max, even east of the 6" line. It's also possible these "lines" could shift 20-30 miles, i.e., the 2" line could be more like along route 1 (a few models are showing a sharper cutoff from 95 and NW of there) - I think that's the biggest bust potential on the table right now.
One other observation is that the storm looks to be a little earlier than thought before, i.e., the heaviest snow for most will be before dawn with lighter snows through 8-9 am. Accumulating snow could be over in Philly by 7 am and by 9 am in NYC, which would lessen rush hour impacts a bit, as visibility might not be the issue it would be if the heavier snow was during rush hour. Still some likely model variations coming up and the actual system will likely have some surprises, as usual.