What inspired this wish list request? There is only one test available in the platform response screening: seems to be ANOVA that gives the pvalue testing the effect of X on response Y.
What is the improvement you would like to see ? It should be great to have non-parametrical test such as Wilcoxon test or other, or the choice of the test to be done in order to get the advantages of controlling the false-discovery rate when multiple tests.
Why is this idea important ? It's important to improve the statistical theory behind "Pvalue" output of this platform which do not give any information on the test done, that seems not adpated to different distibutions of X or Y.