As Alan Case put so well, the author should have multiplied by 100 instead of 1000 and that would have been %strain, which is a meaningful engineering measurement.
% strain at failure is not the best indicator of bow performance at all though because it ignores the force required to achieve the strain and it ignores mass.
IMO the best bow index would be MOR/MOE x MOR/density which is at least roughly proportional to energy storage per unit mass. That formula basically amounts to a rough estimation of the area under the stress strain curve divided by the density.
It's still not perfect because as Alan noted, the wood may have already exceeded the elastic limit at the strain value corresponding to MOR; and more importantly for comparing different woods, some may fail very near the elastic limit, and others may undergo significant plastic deformation before failure. Since this is not necessarily constant across all woods, it makes it very difficult to use this type of data for quantitative or even qualitative comparative analysis.
Ideally you would test a whole bunch of woods using a single protocol and plot and publish stress strain curves for all the samples. Then you could do a pretty good retrospective analysis to determine energy storage/unit mass. Anyone have access to an Instron and a lot of free time. Lol.
I would test all the species in compression, and tension separately to get the most comprehensive data. Three point bending, or four point bending wouldn't be my first choice personally.
Sorry for the technical rant, but that's my take.
