The QQplot figure contains point estimates of the GM, GSD and 95th percentile calculated using the traditionnal equations ( augmented by the regression on order statistics approach when there are NDs ). These values will generally be different from the Bayesian estimates shown above because of ND management and the expostats priors. Although the differences can be explained and should not lead to widely different decisions, they might cause confusion and I would tend to not display them. Any opinion ?
I don’t think there is sufficient value in adding the GM , GSD and 95th%ile on the QQ plot. As you point out is becomes a source of confusion, also the usefulness of the GM in OH statistics as a compliance measure is lower than the parameter estimates. It also creates more clutter in the QQ plot. Leave the QQ plot to just demonstrate linearity of the data.