I-95 forecasting challenge in eastern VA

Alright I just wanted to do a quick update now that the 12z afternoon weather models are in.  I'm not going to go into too much detail with each individual model, but I did want to point out one thing.  One of the biggest forecasting challenges with this storm has been forecasting snow amounts for the I-95 area in Virginia.  The discrepancy over how much areas like Fredericksburg/ Richmond/ Williamsburg/ Emporia see is going to come down to how much of this precipitation falls as snow.   Just wanted to clarify my thoughts and why I see things playing out the way I do.

I-95/Greater Richmond Area/Tidewater Area
I understand the excitement over the past several runs of the nam/gfs that have shown the Richmond area getting absolutely plastered on the snow maps with 12+" of snow.  However, these maps do an awful job distinguishing between what falls as snow and what falls as sleet/freezing rain.  I'm going to use the 18z nam to show what I mean.

Here is the surface map on the left/ 850 mb (near 5,000 feet) temperatures on the right.  Usually the rain-snow line is somewhere near the 0 C isotherm at 850 mb.  This map is for Wedneday evening near onset.  You can see that the 0 C isotherm (in red on the map to the right) is south of all of Virginia.  This suggest that most areas start out as a period of snow.

Now moving forward in time to Thursday morning, you can see that the 0 C isotherm has now pushed west of Richmond, changing much of I-95 and points east over to a mix.  Areas in the tidewater at this point would likely just be a cold rain.

Now looking at sounding data for Richmond (kric) for the same image shown above, you can see that this would likely be sleet if not freezing rain.  The skew T forecast represents different layers of the atmosphere.  Everything to the left of the blue line I highlighted would be below freezing. There's a warm nose around 800-900mb...hence why the 850mb map above shows 850 temps above freezing.  This warm layer would melt any falling snow flake, leading to either sleet or freezing rain depending on the rest of the layer temperatures.

Taking a look at the Euro and the Gfs, both of these models get the 850mb freezing line even further west.  Here's the Euro:

And the Gfs:

Now with all this being said, there is the argument "well I don't think the 850 freezing line gets as far west as that".  Fair enough.  I just don't think you should look at a snow map from one of the runs of the nam and buy 8-12" of snow falling.  Right now I see this line between all snow and a sleet/snow mix setting up somewhere from Danville to Farmville to Fredericksburg.  This would keep heaviest snow fall to the west of this line and decrease it to the east.  Exactly when this changeover occurs will be important, but very tricky to forecast.  Like I mentioned before, the rain line may even try to creep in close to I95.  This leads me to believe that Richmond will be closer to the lower end of 4-8" amounts rather than the 8+ for areas that see more of this precipitation fall as snow farther west.  If the 850 temperatures remain colder longer, the higher side of this 4-8 would work out.