> I've heard that these kind of experiments are impossible to do today, due to "ethics" rules in science.
If you're referring to the fact that the experimenters had to lie to the subjects and deceive them, not just in one individual statement but over the whole experimental scenario, the reason why that's not a good idea is actual practical, not ethical: constructing a false scenario and properly controlling what the subjects believe in a false scenario is very hard.
For example, consider this post by The Last Psychiatrist:
"He didn't stop because of moral courage; he stopped because he thought he was being played."
In other words, he figured out that it was a false scenario and his refusal was based on that, not on any moral objection to shocking people. But he got counted as a "refusal". How many other people who didn't refuse had some intuitive sense that the whole thing was a setup, thereby dampening whatever moral impulses they had? There's no way of knowing. The fact that it was a setup changes the whole experiment, as compared to one where all of the information being given to the subjects is truthful.
That's right. Universities' "institutional review boards" (IRBs) don't permit scholars to tell subjects that they are delivering very heavy shocks. Jerry Burger of Santa Clara University took this point up in his replication article (which was necessarily only a partial replication).
Those rules are of course another example of obeying authorities :)