I found that rather fascinating.
Some immediate questions come to mind, like: at what point do you start tracking hours as sleep deprived versus not? What areas of cognitive performance are being evaluated exactly? How was the research collected and what was the sample size?
Any time I see explicit numbers like this connected to neuroscience a bunch of little red flags start to fly up. But then again, I didn't sleep well last night.
Also "ability to solve complex tasks" seems like it might be difficult to quantify and I can't remember how they did that.
But it's fairly easy to measure the difference in reaction times and capacity to remember strings of numbers.