The very heavy snowfall held out through May slowing the predicted May decline in the 50 day sea ice model. That snow is almost gone in north America now so I think the melt rates are going to start reflecting the model prediction soon. Siberia is warming up fast, but has more snow to melt. For predictive purposes it's probably better to smooth that curve out instead of predicting a cliff, in my opinion. But we all know how hard it is to predict the weather so let's not get overconfident in forecasts. No matter how we look at it, the forecast for the ice looks bad.
The GFS may overcook snow melt, but for what it's worth here's the forecast loss over the next week. It may be too aggressive, but if the forecast heat materialises in Siberia the snow can't last much longer there.
edit: - I added the attachment I'd had forgotten, and the replaced with one that will play /edit
With the state of the ice on that side and a series of lows in the Kara Sea blowing into the Barents and sucking warm moist air in, more and more open water will appear and start warming. Does the ongoing weather setup of a high surrounded by lows cause rotation and compaction of the centre, while lows disperse ice at the edges? With the mobility and fragility and thinness could the pack become entirely separated from the margins of the Arctic Ocean, or even split into pieces rather than develop arms.
If the summer is sunny and the Ocean takes up a large amount of heat, then the autumn turns stormy like last year the minimum could come very late. During the winter on the freezing thread it was suggested(if I can trust my memory) that when we go sub 1million km2 sometime in the next few years, it could happen in October.
It's too early to discount the possibility of going icefree this year. Things have started ominously. We just have to watch what unfolds