I update my probability that the execution of KS tests represents a reliable realization to only 35%, with a confidence interval of +/- 10%. The problem is that the test itself does not validate the quality of input data nor the relevance of the thresholds chosen for evaluation. If the data come from opaque or unverified sources, as was the case with certain entities in the Wirecard scandal, then the statistical similarity is superficial. The practical validity of this similarity always depends on the robustness of the underlying assumptions and the integrity of the information used.