I am assuming she was testing video completely blind, right? Blind test ABX? 🤦♂️😉😂🤷♂️
Even n=1 blind tests are effectively nullified by the (clever*) statement that it takes 45 minutes or so for the effect to be apparent.
45 minutes ruins our ability to compare. I'm trying to keep my 17kCDN electrostatic speakers going despite their 12 panels failing, one by one. Each time one fails, I substitute a pair of boxless transmission line speakers. First disk I grit my teeth. By the third disk I say to myself "I can live with these—why am I repairing those expensive speakers?" And after repair, I say the same kind of thing to myself in reverse "This is so much better" yet by the third disk I can't remember (other than intellectually) what the difference was. Only do I find support for the better speakers in long term usage: do I choose to spend my days listening, or doing something else? Do I find myself staying up late at night for just one more disk? That's why I am going to replace all the panels in those speakers: they keep me listening. In just the same way, it would take months of honest self-reporting to do the same with a device like this. But, read on, if you like.
*clever? Well, certainly. It means you cannot conduct any kind of blind test for yourself, as 45 minutes means no worthwhile comparison. If you want to argue about that I have a solution for you. It will take a few weeks and need a helper. Let me suggest a couple of weeks with the device plugged in, and a couple without it. Record your satisfaction each day. Your helper must not tell you whether it is plugged in or not. Do this for about three cycles both ways, which is six weeks. You must not cheat, so the device must be hidden from accidental view. At the end your scores should be analysed, along with those of (as many as possible) other listeners doing the same thing. An independent examiner will take the results and the information from the helpers which says plugged-in or not-plugged-in for each set of two week scores. Then, and only then, will you know whether you are listening to a real effect, or your own expectations.
Many people will say none of this is necessary. My answer is why not do it—properly, with independent oversight and analysis—and prove me wrong in my suspicions? There's nothing to lose if you are right.