Not beyond the realm of possibilities, actually...

There is always an element of uncertainty with TCs, in particular because of the unknowns in how they will interact with the surrounding atmospheric medium. In this situation, we have cyclic trough and ridge occurrence taking place in the general regions NW-N-NE of Bertha's current direction of motion. We can assess from common wisdom that troughs will tend to excite a northward/northeast motion, but ridges will stunt that and/or even resume west motions. The problem is, there's a spectrum of possibilities in how they will effect Bertha's track (and to some extent intensity) between these two plausibilities, which depends entirely on correctly modeling the strengths and timings of the synoptics as described.

Not an easy task, particularly when subtle permutations can lead to unexpected track shifts, few of which can be correctly assessed beyond a few days time.

In general though, the 12z GFS is not really that off the going thinking as of 11am. The official call is a general NW or even N motion that will be slow, and then a potentially stall 3 days from now. This is essentially what the GFS is indicating as far as I can tell. The question in this run is what takes place beyond D4... It seems the GFS is trying to lesson the latitudes in which that next trough digs off the east coast, and that is why in this run we don't see as much acceleration toward the graveyard as the other runs/models have intimated.