to add to this, it's equivalent to the difference between trying to justify that one experiment (or one class of experiments) is wrong vs several dozens of clas

to add to this, it's equivalent to the difference between trying to justify that one experiment (or one class of experiments) is wrong vs several dozens of classes of thousands of experiments are all subtly wrong so that this one experiment can be right.