Daniel, you seem to be confusing some things. At the moment, both volume and extent may well be decreasing linearly and at different rates - nothing wrong with that. The thickness of the ice is changing which is saying the same thing in a different way.
Of course, both metrics will have to reach zero at the same time (since you can't have one without the other), so if e.g. volume decreases linearly and at a faster rate than extent, and continues to do so until it reaches zero, then at some point extent will have to decrease faster and thus deviate from any earlier linear trend. Because both have to reach zero at the same time.
Other scenarios are of course possible - volume decrease may flatten out to match extent, or both may change their trends. But one thing is absolutely certain: If one metric reaches zero, the other will by necessity reach zero at the same time.
As you have pointed our repeatedly, melting is primarily from above and below. Once melt starts, 1 meter thick ice will lose volume at the same rate as 2 m thick ice.
Each winter, ice in the Arctic reaches very similar extent, constrained by geography - the Arctic Ocean simply freezes over, with small exceptions. So each melting season starts with similar extent, but since volume is going down, the average thickness is less each year.
Average thickness at maximum has been going down from some 2.6 meters in the 80's to around 1.7 meters the last few years (from eyeballing the graphs - not a scientific number!), a drop of some 35%. Extent at max has gone down from some 15.5 million km2 in the 80's to just under 14 now, a drop of 10%. So thickness at max is going down some three times faster than extent at max.
Each year the thickness drops some 0.9 meters between max and min, and this has not really changed during the last 40 years. If we take this to mean that a given thickness of ice loses 0.9 meters during melt season, any ice thicker than that should survive the summer, the rest melts out (again the 0.9 figure is from eyeballing the graph - some years, e.g. 2017 lost only 0.7 meters, admittedly from a very low start, while 2012 lost 0.9 meters in a record-breaking year).
So thickness at max (and hence volume at max) seems to be the important variable here - extent is more or less the same each winter, and the rate of loss of thickness is the same as well. So extent at minimum is perhaps driven mostly by thickness at maximum?
It's interesting to note that the decline in extent at minimum is 45% (from 7.2 in the 80's to some 4 million km2 now) which seems to coincide surprisingly well between a 10% fall en extent at max + 35% fall in thickness at max.
But this is only playing around with numbers. My gut feeling is the same as has been expressed by others here - as thickness keeps going down, a rapid melt year could simply make it all more or less vanish. So both metrics at minimum, volume and extent, will deviate sharply downwards form any linear trend, getting damn close to zero (and well below 1million km2 or 1000 km3 in one big leap). But given a linear trend in thickness decline, it may well take another 20 years before thickness is below 1 meter on average at max.