I don't know the details of how wheels are load rated, but I do know a bit about International Engineering Standards, and Material science.
I would guess that the mode failures could be various, but sudden catastrophic collapse is probably the least likely to occur.
Usually the standards are more about good serviceability; which is breached way before total collapse.
When a wheel comes under undue stress, high elastic deformation (i.e. it temporarily stretches) or plastic deformation(it permanently bends/buckles) are going to cause all sorts of problems with tracking/ balancing/tyre inflation poor ride and control/ undue vibration and stress either altering ride comport/stability/brake performance or unduly stressing other components in the drive line/suspension.
The standards will also allow for time related corrosion/weathering/materials becoming brittle/stress fractures over time of a component etc etc.
The failure mechanism will invariably be mounting a kerb or hitting a pot hole etc etc; when
the dynamic loads are probably 10 times or more than general driving. A huge factor in these sudden dynamic impact loads is the forces generated by sudden deceleration or impact, and very low profile tyres are far far worse than standard profile tyres at allowing this stress to occur.
The wheel failure we all know about, are a friend down the pub, whose very low profile wheel/tyre combo being buckled/cracked/rim damage causing sudden deflation etc etc when kerbing or pott-holing. I think a slightly under spec loaded rated wheel with a 125mm of side wall of rubber; is about 100 times less likely to fail than a correctly load rated wheel with only 60mm of shocking absorbing side wall, given the state of todays roads.
Not the law, just good old fashioned engineering imo