I love these sorts of articles. They criticize Milgram for a lack of scientific integrity, but nobody seems to point out that prisons aren't a bastion of scientific integrity, either. If anything, Milgram's sloppiness made his study a better model of the institutions in question. He allowed his human biases to overcome his duty -- which, as his own results show, is exactly how these things happen.
Milgram's experiment wasn't about prisons (at least not directly), and it didn't really seek to model any particular situation. It just sought to examine people's response to authority, and the setting chosen was merely a convenient way to do that.
If you're referring to the Stanford Prison Experiment, there's no point making something a "better" model if you lose your controlled environment in the process. The fact that Zimbardo interfered might be an interesting observation, but since it's not a repeatable experiment, it becomes merely a well documented anecdote.
Also, this isn't what the articles were about. They're mostly just presenting alternative explanations for the results, and the experimental evidence which supports them.
