Exactly. Which renders the logic behind the experiment false. But since the insight was true, it eventually led to experiments that yielded true results. (Notice I didn't say the insight was false, I supported your idea by emphasizing that even false insights can be useful.)