Tealight than you again for your amazing cryosphere computing resources. I keep wondering how the extra added insolation, the AWP anomaly, affects the melting season. It is certain that extra AWP in the inner Arctic basin has a significant effect. 2016 was a great example, where the extra AWP in April and May was the background that enabled the GAC in August, and also delayed the freezing season significantly, so that winter 2017 saw the lowest volume by far on record. All this despite the cloudy weather in June and July 2016 that prevented a new September extent record.
I am also quite confident that AWP anomalies in connected peripheral seas, such as the Bering and the Barents, affect the melting season progress in the central Arctic, where the minimum is set. However, I believe AWP anomalies in the following seas do not have much effect on September minima, though they do contribute to general AGW:
Hudson Bay - far and geographically disconnected from the central arctic, melts out anyway every summer.
Sea of Okhotsk - same as above.
Gulf of St. Lawrence - same as above.
Baffin Bay - downstream (in terms of main currents) of the CAA and central Arctic, melts anyway every summer. Its gained heat flows south.
Greenland Sea - downstream of the central Arctic, its ice in September is ice that was exported from the central Arctic late in the melting season. Its gained heat flows south.
I personally would be happy to see a "central basin and connected seas" cumulative AWP anomaly graph, excluding the seas listed above, in addition to the pan-arctic graph.