Go's decision here also defeats the next lazy thing people do, which is they run it once, and then they copy-paste whatever the results are as "correct" into their test case. With the randomisation their next run is also randomised and this "correct" answer will now be wrong so their tests fail again.

As a result most people will go "Oh, I guess I should actually test the outcome I care about" (good) or "Oh, testing is hard, I'll skip it" (bad, but at least not a problem for the team maintaining the standard library). Very few will say to themselves "I bet I can defeat the randomisation somehow with a different incorrect test design" because humans are lazy.