It's the nature of psychology. You can't tell someone what you are trying to study, or else you will always get confirmation bias.
These experiments, they rely on human nature, for people to act with them, for them to work. They are dangerous, yes, but we have to do them to really learn about people. So long as the subjects effected are volunteers for what they are doing, and all effects to them are due to the way the person receives what they are told, I think that the odd experiments mentioned in the article are OK. They are invaluable for learning what it means to be human.
Crossing a line is putting people in pain by shocking them, not the deceit. I don't have much sympathy for those who believed and formed a religion, especially because those people stand as a shining example of why we need to be cautious about such topics today.