We ran multiple tests to answer this question ourselves. The graph on the left is showing you the statistical precision we achieve with a given amount of testers. You can see that as soon as we reach 20 testers, we reach ~80% precision. That doesn’t mean we can’t have more testers but to improve the precision significantly on that level you’d have to double your testers.
That’s why we generally recommend 20 testers and multiple tests after each other.