This is why LR Op forecasts need to be taken with a grain of salt. If you look at the 0z 2/19 run and compare it with 6z 2/21, the differences between modeled snow and nada are subtle:
https://www.tropicaltidbits.com/analysis/models/?model=gfs®ion=us&pkg=mslp_pcpn_frzn&runtime=2023021900&fh=162
https://www.tropicaltidbits.com/analysis/models/?model=gfs®ion=us&pkg=mslp_pcpn_frzn&runtime=2023022106&fh=108
One model had more of a return flow, I guess, with higher PWATs and I suppose a better thermal gradient. I'm not really sure, but at least for a setup that doesn't involve a well-defined mid-level vort, it's probably best not to get too excited over seeing blues this far out.