All the limitations you mention are already explicit in the text.

You have any other studies demonstrating lack of cognitive decline in 22 year olds on 4 hours sleep? Otherwise, you're just poo pooing on someone having actually having done the damn experiment. Yeah, n=1, but it's better than your n=0.