> Tensile-test a strip of iron and a strip of copper and you'll find
> they're more similar than dissimilar.

It's a matter of perspective. I reload ammunition. I can reload brass and soft metal cases (which copper is considered to be), but not steel cases. Ferous metal is far to hard and unyielding to work with. The differences between the materials are vast. But compare them to chewing gum, and they then appear to be quite similar.

> Right, failure modes differ and have to be considered separately. Landing
> on a flat rock will result in bending stresses in the material, landing on
> a pointed rock will lead to a combination of both bending and shear
> stresses.

Forget not the radius of that bend. This whole area right here is a source of mutual disagreement between just about every boat builder, operator and engineer. What factors are important, what are not, etc. No one answer exists, there is no "optimum" design. It's an area of engineering, designing, and building that is argued regularly, and frequently, in this web board.

> Not necessarily. The yielding may only result in a white spot where the
> cloth separates from the wood and the wood underneath is crushed, but the
> kayak is still intact and readily paddled. Failure is when your hull has
> been punctured or fractured and water enters. I would rather paddle a
> kayak with a dent than a kayak with a bow hatch full of water.

I'll go along with that.

> You can make it stronger by making it less stiff throughout.

Yes. You possibly can, but not inherently can, it's a target easily missed, and it may not even be desired. Especially if poor modeling is used. Such as flat plate analysis for a compound curved surface. All sorts of mistakes would result from that.