My thoughts on explaining A vs B on 1 (hopefully this makes sense): A is computerized models behave like real world counterparts=theory probably true. This leaves open the possibility that model doesn't behave like real world counterparts=could also be true. But does that matter? No, because we know from the stimulus that they definitely do behave like real world. Versus B is if theory is correct=models behave like real world. This leaves open the possibility that if theory is not correct=models could also behave like real world. Does this matter? Yes, because it means our model behaving like real world doesn't actually prove the theory is true. This is how I think about it (I get hung up on the other possibilities still being on the table, but deciding if they actually matter is useful) so hope it can help someone else.