Jawge and Pearl, you both have valid points: Wood is 3 to 4 times stronger in tension than compression, but there is a point where a the back of a sapling is so highly crowned that less than a third of it's back is doing the tension work.
However, the few sapling bows I have had fail all failed by chrysalling, meaning the back was stronger than the belly.
Pearl, the narrow crown, or the narrowed back of a trapazoidal cross section turns out to help performance by reducing limb weight--in tension, wood has almost no stretch before breaking. So, it doesn't hold that we are balancing the elasticity of back and belly (as I believed until shown otherwise in testing literature). We just don't need as much wood on the back to be as strong as the belly.
Jim Davis