I think it's important to walk before you run. If social scientists aren't able to get compatible results from the same experiment done by two different teams,

I think it's important to walk before you run. If social scientists aren't able to get compatible results from the same experiment done by two different teams, then I seriously doubt their ability to extract any relevant subtle truths about human behavior. The c. elegans story you linked is a great example of this. Getting the little details right matters. I can't imagine having any luck doing that while constantly changing the big details.