A couple of my half-baked thoughts about the modeling in general:
Good job for the most part with QPF (not withstanding the bizarre ukmet qpf hole and some other model run faux pas)
Decent job with the sleet extent once they caught on to the warm nose.
I thought I had a good handle from the modeling on what was going to happen in this region, and that was correct. I had a lot less confidence in what was going to happen in the Mid-Atlantic region. Seemed like more variation in the modeling for there, although I didn't look at it as closely.
Most modeling held off too long on flipping LI back to snow last night. I had the impression we'd wake up to some freezing drizzle and temps between 25 and 30; 18⁰ and snowing is what verified. At least on the north shore we were going back to snow around 10pm. My loosely considered model consensus had a substantial warm nose persisting through most of the overnight.
Overall for my insulated little corner of the world, my landscape looks very much like what I was expecting it to a few days ago...moreso than usual so subjectively, the modeling was very good.
The NWS did a good job too. They more or less stayed with a forecast that eventually verified well through many forecast cycles. Complaining about tweaks like adjusting P&C totals from 9 to 14 inches down to 7 to 13 inches is petty, especially when both pretty well reflected the eventual outcome.