zero confidence: low, but higher than any other estimate. For this year to loose enough volume to bring sea ice volume to zero is consistent with an exponential trend in volume declines from October to September. It would not be consistent however with the extent trend. But the extent trend isn't likely to persist to zero from what I understand. Also, there seem to be more clouds so far in May compared to last year, judging by a look at the infrared satellite view. According to the new study published in nature by Kapsch (http://www.nature.com/nclimate/journal/vaop/ncurrent/full/nclimate1884.html), low-ice-years (i.e. low minimums) are correlated with high cloud cover in springtime through middle May. And 2012 was itself a low-ice-year. So if this year continues to have more cloud cover through May, it would stand to reason that melt this year would at least exceed the trend. For the ice to reach zero, about 18% additional volume would have to be lost between now and minimum, compared to 2012. That's a few thousand cu. km of extra ice. If there are more open areas of ocean, the increased mixing and wave action will break up and melt the ice more efficiently. Plus, beneath a certain point, the ice will probably melt much faster than ever recorded as it breaks up into many small pieces with a much higher exposed surface area/volume. Based on the volume trend of previous years, it appears we could rapidly be approaching that situation in 2013.

If you haven't seen the episode, you should probably know that the giant trash cans in their van are full of gasoline. The plot summary from tv.com is "To capitalize on the high gas prices, the guys invest in barrels of gasoline that they plan to sell door to door. Meanwhile, Dee and Frank resort to some drastic measures to stop Bruce Mathis from donating Barbara's money to a Muslim cultural center." http://www.tv.com/shows/its-always-sunny-in-philadelphia/the-gang-solves-the-gas-crisis-1228275/

...and the early melt could also indicate warmer incoming ocean currents, a trend that might last for several months naturally. Or because of global warming, perhaps essentially forever, with ever increasing temperature of the water. In which case, the absolute volume of ice lost each year from max to min should keep increasing, if everything else remains constant (which it doesn't obviously, but we know that and can work around it).

...to elaborate on the graph I just linked to, the basic assumption underlying the trend is that more days of melt mean more total melt. I estimated day 105 for max volume based on tracking of the NAVY CICE model. I know their forecasts aren't wonderful, but their hindcasts seem alright. Plus, the area per pixel is scaled in such a way to make interpreting results by pixel color count possible after splitting the thicknesses and using a histogram ;)

I've plotted sea-ice volume loss between March and Sept for all PIOMAS years versus date of max volume. There looks to be a relationship. 2013 numbers to reach ice-free state included: http://www.fairfaxclimatewatch.com/blog/2013/04/arctic-sea-ice-breaking-up-holes-forming.html