If the material composition of the highback was altered slightly, and tests were done as described above, then the 3% sounds like it could just be a number the marketing team got ahold of.
Unless, combined with other changes in re:flex or whatnot, the sum of the small changes would be somehow noticeable.
I bet the margin of error for the stiffness of run of mass produced highback pieces is at least 1.5%.
Exactly. The 3% number is complete bullshit. What specific aspect was measured that changed by 3% percent?
The force required to break the highback? Which would be completely useless information...
Or the force to deflect/twist the highback by a certain degree/extent? In that case how about a different degree/extent - I guarantee you it will not be 3% and could be 10% more or even 10% less.
It is bad enough that the industry is trying to rate flex for boards and bindings on a single number scale, but at least those numbers can offer some guidance. Stating that a highback is 3% stiffer is just complete and utter lunacy and completely meaningless (as pointed out by others already, even the manufacturing tolerances are more than that).