I posted my experience with jplay in another thread after looking at its code and trying it with Foobar2000, and I just happened to click on this thread.
I don't think it's a coincidence that Chuck Zeilig, one of the authors of this seriously flawed study is also a prominent advocate of jplay on the jplay website. I loved it when he stated (I'm paraphrasing) "if you can't repeat our results, then you've done something wrong" - revealing that he's completely aware of the need for repeatability for valid scientific tests, and he decided to preempt and mock calls for repeatability. That's just a small example of the lack fo compliance with basic, but vital, test standards for any valid test.
Chuck's involvement in jplay as an advocate and as an author of the TAS article stinks of conflict of interest, at best.