The Kara/Northern Beaufort dipole is bad for the ice. The actual melt processes can be very complex especially at this time of year and the last few posts have highlighted various elements at play.
We've got sun/ocean/rain/snow/ice all at play. But it is far too simplistic to look at just one measure (such as 850hPa temperatures) and infer that subsidence will melt the ice. That's not how it works.
The ESRL expermiental sea ice pages drill down far more detail. It takes considerable time to tease through and assimilate the wealth of forecasts on that page.
I've had a look through their forecasts on precipitation, top melt, bottom melt and albedo changes over the next 5 or 6 days.
First image below is the top melt at +42. At this time the low over the Laptev increases top melt with wind and most likely, rain.
Next image is of top melt at +126. Now we see a line of considerable top melt spreading up through the centre of the Arctic.
This area of top melt appears just after a belt of precipitation passes through, presumably rain/sleet. This is the boundary between the dipole. A classic frontal zone.
The ice under the centre of the high is showing very little or no top melt.
Meanwhile underneath, bottom melt is slower to change but is slowly spreading into the centre as the arctic ocean is still warming and will be probably for another 2 months. Bottom melt increases especially when onshore (to the ice edge) winds drive the warming open water currents under the ice. Persistent winds attack the Beaufort edge and bottom melt peaks there at about +132 (next image). Later in the forecast winds swing back more northerly in that region as the high moves over towards the centre of the Arctic and bottom melt decreases.
https://www.esrl.noaa.gov/psd/forecasts/seaice/