The problem there is the subjectivity of the test. Some people will be easily convinced, others not at all.

If we are testing for the objective existence of something that’s either there or not regardless of who’s performing the test, then there will be objective criteria for what makes for a good or a bad version of the test.