Oh you're right that is a misread on my part, the appendix charts don't say that. I think they're just useless then though? Since they're reporting absolute error (on a log10 scale) we can't assess the relative to compare to the 'within an order of magnitude' claim in the text.