They can be in the right context but not when they're comparing interventions which cannot be blinded. And where one of the interventions changes something else (in this case, the amount of other foods you eat if you have to eat two kiwi fruit instead of just popping a pill).