And what makes you think that they didn't account for the "I don't knows"?
It generally surprises me how, from reading an 100 high mile description of an article, some commenters always assume (without any evidence) that the results are due to some extremely naive methodological mistake.
If you're teaching a subject to someone, you will often come across situations where what is lacking is not knowledge, but confidence or willingness to apply the knowledge. I see that often in my son, and I've seen that at work: People say they don't know, or make a crap guess even in situations where I know that they know the actual answer. Often some prodding or "confidence boosting" will make them produce the real answer very quickly, and often subsequent answers appears to be forthcoming a lot quicker.
It'd be very surprising to me if you can't systematically improve peoples response by increasing their confidence in their ability to answer. The more interesting question to me is by how much, and with how little encouragement.
I therefore make the very personal and questionable choice to ignore (most of) them and go straight to the neuroscience people - which in general have a way higher understanding and higher scientific standards.
Since the article doesn't presume to explain its results, the presence or absence of mistakes is moot. Science isn't about descriptions -- that's metrology. The threshold of science is crossed when someone dares to offer a testable, falsifiable explanation. But in psychology, that rarely happens, and psychology has no central defining theories (explanations) such as are found in scientific fields.
Well, yes, if someone expects you to know the answer it seems pretty straight forward that you would be more inclined to make a guess rather than saying "I don't know." All of us who ever watched a quiz show on TV know that people who guess an answer instead of passing will sometimes get it right. More often than people who pass, actually.