Given the many replications of variations of Milgram's experiments, any flaws in his original experiment don't detract from the overall result.

See https://behavioralscientist.org/how-would-people-behave-in-m... for some of that replication data.