August 9 is about the date that 60 deg N and 90 deg N receive the same amount of insolation.
And it is still pretty high (the same as the equator receives at this date) :
Now, with some 400 W/m^2 insolation, some 200 W/m^2 will make it to the surface with average weather. Count in a -70 W/m^2 LW (IR) radiation imbalance (see Andreas numbers), this mean that any surface with an albedo higher than 1 - 70/200 = 0.65 will start cooling, while any surface with an albedo lower than that is still warming up.
In other words, small shallow melting ponds will start freezing over, while deep, dark melting ponds will still continue to melt. Pretty much what we see on Obuoy's web cam :
Following the same calculations, open ocean is still absorbing some 130 W/m^2 even with average weather. In clear weather, this can easily top 300 W/m^2. That is still enough power to bottom-melt some 3-8 cm of ice of equal size to the open water.
This means that any ice close to open water is subject to significant bottom-melt from heat absorbed in the nearby open ocean water. With the highly fragmented, already thinned-out, ice this year in the very large ESS-Beaufort-Laptev triangle, means that there will be a LOT of ice melting to come (with more "flash-melts" to come).
The open ocean areas in that triangle is simply still absorbing insane amounts of heat right now.
And thus, I do NOT expect the current rate of ice loss to slow down any time soon (at least not until September).