Is there a purpose served in further tracking of this obscure, obsolete, low resolution rubbish? The size and quality of pixels NSIDC using here is outdated and inappropriate (see wayne's ASIB comments) to the intrinsic size of ice features such as leads.
Sacrificing accurate assessment of the current situation for the sake of long-term consistency (downward trend: are there any convince-able persons left the outside world?), surely that is a backwards-looking agenda.
Below, the boundary between thick and thin ice (green) as determined by U Bremen SMOS is overlain on high resolution sea ice concentration (bluish white) U Hamburg AMSR2. Ice thicker than a half meter is given a gray overtint.
While numerous satellite products are providing a nearly identical edge between open water and ice (with the exception of Hycom thickness which seems to include slurry, grease ice and nilas), only SMOS seems to be accurately tracking growth of new ice oceanwide. Here UB's five grades of intermediate thickness are consolidated to define the position of the 0.5m thickness boundary.
In past years including 2012, that line would have reached the coast by now. In the past, new ice has been capable of ~2 m of growth with 0.5 m only representing roughly a quarter of the season. This year, the line is going nowhere. The time series below shows some back-and-forth but provides no observational support for significant or consistent region-wide thickening of ice during the last 36 days in the Chukchi, Bering Strait and Beaufort.
FishedOut summarized this in a nutshell: a big El Nino burped out heat stored in the ocean since the previous one, leading to too much warm water vapor being advected both north and south (wip's graph), the Arctic is piling on after-effects of its own, making the air and water at its two portals too warm, and this together with the long-term trend proved enough to tip over the climate. That makes for a good default scenario; we'll soon see if the coming melt season continues to be consistent with it.