I tried to make a point a few days ago, and may have miscommunicated it, so I hope you don't mind if I try again, only with some simplification--and some new art.
That point was that over the years, total annual volume loss has remained fairly constant, with a yearly average of between 16.5k and 17.0k km3 disappearing from maximum to minimum. The "death spiral" is, thus, not due to increasingly greater melt, but is in fact almost entirely the product of decreasing annual maxima.
That being said, then, I threw together this dual plot showing volume losses from each year's maximum through June 30, and from July 1 through each year's minimum. And it's plain to see that, as the years have gone by, there's increasingly been more of the former and less of the latter. In fact, because of that c. 16.75k km3 average, there exists in most cases a loose inverse relationship: more early melt means less later melt, and vice versa.
Now, the 2013 volume maximum was the lowest on record yet (though admittedly not by much). And between that maximum and June 30, the amount of ice lost was less than it's been for a number of years. My estimation, then, is that the post-June loss this year will be greater than it's been for a number of years. And because of that, the 2013 volume minimum will be lower than some are predicting. Again, I'm not necessarily saying we'll see a new record (although that's certainly well within the bounds of possibility). I'm just saying that this year will show once again that any "recovery" is purely imaginary.