Um. That article describes a study where the researchers took recordings of different violins, looked for a difference in the strads, and found one. That doesn'

Um. That article describes a study where the researchers took recordings of different violins, looked for a difference in the strads, and found one. That doesn't establish that strads sound different, it presupposes it. Surely if they wanted to show that strads sound different they'd need to test whether their method lets them predict whether an unknown recording is a strad or not?