I don't get that. If the temps and qpf are means of all members, then so is the snow a mean of all members. If the mean temps for an area are low to mid 20's, and upper level means are well below freezing for the same area, shouldn't the mean snow for that area correspond with those indicators?
Otherwise, it would appear that the cutting members are skewing the snow totals. The placement of the low on the GEFS is also a mean, and the resulting metrics should be based on that.
I think the problem with most of the ensemble models is that they take a mean of the snow totals of all members (including the cutters showing little or zero) and plug that into their maps, when in reality the maps should be reflecting the snow resulting from the placing of the mean slp and resultant lower and upper level means. The storm vista maps appear to take the latter approach with the GEFS.