Two reasons I believe.
1) The models seem to struggle with this scenario, not sure why, but they just do not pick up on this. Perhaps because it is a relatively narrow area that gets hit and primarily due to elevation.
2) From the human side, just like anything else in life, the more we see it or experience it, the better we get at it (usually!). Since this is such a rare event (happening once or twice a year on average), we do not get as much of a chance to try and hone our forecasting skills with it.
-John
