I think it'd be cool to get a bunch of statisticians together to review Optimizely. Print out pages of the UI, remove the logo, and then ask them to write down opinions on the validity of the results UI they're shown.
The problem is that it's a lot of work and there's really no incentive to do so. At the end I worry the group could be retaliated against and/or be considered complainers. There really isn't any benefit to doing a review like this that I could see. Plus if you're a statistician like me, you already know the conclusion, a kind of blind study would only build credibility for non-statisticians. :-/
https://news.ycombinator.com/item?id=10873226