All reviewers commented on that it would have been interesting to see results on real-world datasets

We are very happy to hear that all reviewers found the paper interesting and well written. Below we provide point-by-point responses to selected comments. All reviewers commented on that it would have been interesting to see results on real-world datasets. "controlled way" results in a cleaner and more informative numerical evaluation. The distribution-free bounds are typically between 0.99 and 1, We will add these additional experimental results to the revised supplementary material.