Another way of looking at stability -

A boat's primary stability is determined by its metacentric height GM, which is the vertical difference between the center of gravity G and the metacenter M. As small heel angles the boat "rotates" about M; think of it as a pendulum. The center of gravity is fairly straightforward but metacenter is not an easy thing to define simply; but here goes:

The center of buoyancy B is the center of gravity of the volume of water displaced by the hull obviously it's below the waterline. As the hull heels a small amount B moves laterally, a vertical line through B will intersect the centerline at the metacenter M. The movement of B off-center can be calculated manually for a small angle of heel, but it's a whole lot easier to have a boat design software package provide it in the form of a value for KM, the hieght of the metacenter above the keel.

That's an over-simplification unfortunately. For a kayak the center of gravity G is determined mostly by the center of gravity of the paddler(s), the boat making a minor contribution apart from seat height, and the paddler's physical characteristics are beyond the designer's control. Also the metacenter doesn't stay put because part of the hull that was above the waterline is now below it, so the location of M varies with heel angle - the variation with heel angle determines secondary stability, which can also be affected if the paddler shifts laterally. Just to finish it off, M also varies with load - it usually gets lower as load is increased. Great news for the larger paddler . . .

It's a reasonable assumption that increasing waterline beam will raise the metacenter M, but it also tends to raise the center of gravity G, so GM - the real measure of primary stability - doesn't increase as quickly as you might expect. Sure you can toss in oodles of rocker - screwing the tracking. Sigh! This is why boat designers get the big bucks. Well, they should anyway . . .