Odd that comments complain it was "too easy."

This is science. They're disproving a very narrow, specific hypothesis about a particular algorithm. They're not trying to disprove deepfakes in general. It's just one experiment collecting data on one setup.