This is the same exercise I have been doing for the last three days – since before NSIDC changed their data. I overlaid the current DMI graph (black) on the original version of the April 16 NSIDC graph (blue – no longer available online.) It shows that ice extent has not declined for more than a week, and has now crossed the 1979-2000 mean.
Further confirmation of this is that yesterday’s NSIDC map showed more area of excess ice than missing ice.
NSIDC says that they have introduced a two day lag into their graph. Bottom line – if NSIDC doesn’t cross the mean in the next two days – their graph is wrong.