That's not a bad idea Richard, although Stuart is right. The stability of a ski & paddler combination is very dependent upon the height of the Centre of Gravity of the paddler. If we assume its near the belly button of the average person, then yes, most ski's are near neutral, slightly stable, or slightly unstable depending upon design.
However, in terms of a measure as you suggest, this could be done I think. Instead of dumping 80-90 kg of weight directly into the seat, the boat could be supported in a rig that applies a specified load and allows the boat to roll about a specific point (quite low down, simulating a low CoG). One could then measure the righting moment over a range of angles. Obviously this would be a relative measure, against a 'standard' or benchmark case, but potentially quite useful I think.