Thanks to Julienne for explaining what was going on around the periphery.
My questions have to do more with the interior. I made an animation (above) of one region – showing July/September normalised (i.e. adjusted for drift.) I think this makes my observation a little clearer.
Note that the thick multi-year ice (green) has almost all disappeared, while the thinner pink and blue ice from July is essentially untouched. There is no physical model I can think of where this kind of behaviour can happen.
I may be wrong, but it looks to me like something is amiss with either the July map or the September map.