You are absolutely right. The plot actually corresponds to the log of [X]^n/(Kd^n+[X]^n), which is the form used by modellers (both authors are), while the equation in text corresponds the phenomenological curve fitting, [X]^n/(K+[X]^n) where K is an apparent dissociation constant (or an EC50). The easiest path is to correct the text. I will do that in Wikipedia. I am not sure how to do it in the PLoS article.