Not perfect matches, obviously, but that negative tilt over the southern plains to a bowling ball over DC is a good match for Jan. 2011 commuteaggedon.
Yeah, we probably lose the column at 4z or 5z. Not a lot of difference in freezing lines between surface, 925, 850, and 700 is not a problem, so if there is a mix, it would be a short period.
I guess I'm not seeing what you are seeing flipping back and forth between runs. If I were placing the 500 low I'd be inching it even a bit further west from 06z, at least by the time it gets to Cleveland. The big "win" on the 12z run, however, is that the GFS did juice up the front end. This is all about what we can get before the screaming 850 jet (85kts!) gets us by 00z Monday.
I think we all need to be realistic about the situation post-storm also. It doesn't seem to matter which model, but the cold air does not rush in quickly behind this storm. If it does snow along 95, it is going to drip-drip-drip on Monday. This is a storm where we enjoy the front end and then move on to looking for the next one.
Where was the 500mb low passage on the 06z Euro? We desperately need to keep that south of us like the 00z Euro and 06z ICON have. The GFS/CMC solutions are bummers.
You are super wrong about the current forecast models, but you do have some support on the analog side. There are groups out there that are experimenting with machine learning using a trained data set (the analogs). It’ll be interesting to see how that works out.