There is a slight error in the shifting of the curves, but no matter. The point is that if the curves are identical in terms of speed as viewed on these curves, then if they print identically, then the speeds are the same. This is one of the faults in measurements of ISO like this.