Two weeks I wouldn't have a problem with, until the middle of December, sorta like how the EPacific season starts in the middle of the month of May. The relative frequencies of storms in each basin during the last half of May and first half of December are probably about equal as well, at least for the past 10-15 years. It'd only cover a few storms, I believe, but I think is more feasible and, to the general public, less confusing than having a season span two years. Sure, the calendar is merely a construct we use on a daily basis and could be changed if we really wanted to, but there's no real incentive to do that. Just IMO.