You don't need a controlled experiment if you have a good enough understanding of the mechanism, such as with parachutes. But since they apparently had no idea

You don't need a controlled experiment if you have a good enough understanding of the mechanism, such as with parachutes. But since they apparently had no idea how peanut allergies worked nor had any adequate studies, they should have just shrugged their shoulders when asked for advice. Even with parachutes, you could do a study (not a RCT) by looking at historical cases of people falling with and without parachutes. The effect would be so strong that you wouldn't need those clever statistical tricks to tease it out.