The reviewer is incorrect – while a confidence interval of an R2 can be calculated analytically, the appropriate metric would be a confidence interval on the median R2 and Spearman \(\rho\) values that we use for comparison and discussion. Since the distributions are unknown and clearly non-Gaussian (e.g. Figure 2) there are no analytical formulas.
In the revised manuscript, we have added a discussion of calculating confidence intervals using bootstrap methods and have added them to appropriate figures and tables. As usual, our code is provided as supporting information in our GitHub repository.