I wouldn't be the best person to answer this, but I think people that remember this case say that the models from back then did a poor job of picking up convection or something that contributed and by the time that it started adjusting back, it was too late for it to not be a huge surprise.
What I do know is models have progressed so much since then that there are less huge surprises. Even cases like December 2010, there was a hint that a winter storm was possible and it just needed the phase to pan out.
Edit: Now I'm looking into this, and it looks like the computer models had a hard time handling a low pressure system in Canada, along with the system that produced the Carolina Crusher.