FWIW - The edited chart below shows how an NSIDC maximum Extent would compare to other years if refreeze from Oct 28 followed the highest rate in most recent 13 years until reaching a maximum in March 2021. Based on JAXA values posted by gerontocrat:
https://forum.arctic-sea-ice.net/index.php/topic,2975.msg291409.html#msg291409 To create the hypothetical NSIDC max I increased the extrapolated JAXA max by 0.41 M km2 to account for the difference between JAXA and NSIDC record lows. (The extrapolated JAXA low would be 13.49).
Extent value is an incomplete metric to represent the complexity of ASI condition. Additional dimensions (Area, Volume) and qualitative characteristics (Thickness, salinity, mechanical strength, snow cover, temperature?, density?) are missing from a simple measure of maximum Extent. In addition to ice measures, it seems that emerging changes in the Arctic Ocean (open water, wave action, water temperature, thermo-halocline stability, Atlantification, currents, storm potential, less ice pack cohesion with greater floe mobility, and more) are almost all on the side of working against ice retention, not promoting it (a possible exception being jet stream changes described by Francis and Wu 2020,
https://forum.arctic-sea-ice.net/index.php/topic,2692.msg291501.html#msg291501)
Thus, even if the resulting maximum Extent is closer to (or above) the trend line, the 2021 maximum Extent value, whatever it is, will be for ice that on average is almost certainly less resistant to melt than even the most recent historical norm as it heads into the 2021 melt season.
It must also be noted that when the long term trend is extracted, the annual maximum Extent has almost no predictive power for the subsequent September minimum. A dramatically low March maximum at the beginning of the melt season does not tell us what to expect at the end of the melt season.
But with the expected effects of late refreeze on ice quality and melt resistance, I wonder if that previous lack of correlation between preceding maximum to the subsequent minimum will continue. Compounding effects of qualitative decline may emerge from the noise of year-to-year variability to become a separate and identifiable (and measurable, monitored, and reported?) influence. Just guesswork inspired by hypotheticals.