At that forecast length, synoptic scale uncertainty can still be an issue. As I've said several times before now, there's really no point in getting worked up over the details of any specific forecast at this range. Check the ensembles or d(prog)/dt for uncertainty estimation. The 12Z GFS is dramatically different from recent runs. A similar shift or noise can also be seen in recent FIM runs. The 12Z GEFS has noticeable variability in the phase and shape of the trough, and the control member (what you saw) is towards the edge of the GEFS distribution as far as placement of the trough. The 00Z Canadian ensemble is all over the place with that trough. Some members have a "northern" trough just barely crossing onshore with the height min in the Pacific northwest. Some members have a closed low skirting the Mexico-California border. Others have the more favorable placement over the high plains, and yet I see one other member with only broad low heights over the southwest US with a cutoff or extremely negatively tilted trough over the upper Midwest!
I understand you need to plan ahead, but this is the game. Risk and uncertainty are part of it.