That is very true. However, you do have much better reason to believe that the option is neutral than you have to believe that the option is beneficial. In an e

That is very true. However, you do have much better reason to believe that the option is neutral than you have to believe that the option is beneficial. In an example like the one given in the article, you also likely have enough statistical power to be reasonably confident that the option is close to neutral, so if you're making a negative decision, it's probably not strongly negative.