Hmmm. I think his scoring algorithm is a bit flawed. The first time I got distracted and missed one at a high level, then did quite well after, and ended up with a score of like 10.5hz. A proper algorithm would boot any result that was obviously a "flyer" as we called them when I was shooting.
The second time I didn't miss anything except one in the 1.5hz set and several (about 1 out of 5, maybe) in the .75hz set - yet this scored me at 2.4hz.
In fact, the second time through I made it through several of the 0.75hz samples before missing one. When I missed one, it started going back up, then back down again, and did that about three times. Each time I got the first few samples in the .75hz set correct and it was on the second or third iteration that I missed one in the 1.5hz range.
I just don't see how a single miss at 1.5hz and maybe 20 or 30% misses (at the outside) at .75hz can give a score of 2.4hz.
Anyway, pretty cool little test but I'd take the final score with a large dose of salt. Also, not sure of the usefulness of it. I scored pretty poorly (about 8 cents) yet I know that in a quiet room I can tune by ear (referencing string to string) about as tightly as my strobe tuner can measure - i.e. well under a cent. For tuning, once you get within a few cents you're listening for the beat note, not for whether one note is higher or lower than the other (this is why you always start tuning from below your reference string so you know you need to tune up until you get the beat note - after that it's just a matter of matching frequencies so the beat note goes away).
The important test for musicians is whether you can tell if the note you are playing is in tune with the note someone else is playing - and that's best done with the notes being played at the same time.
Still, fun.
Edited to add: Oh, and FYI i was using an $18 set of Philips earbuds and the grandkids were screaming in the other room...
John