You are misunderstanding the article. There is no such thing as “fooling a null test”. It is a clear objective method to audibly determine the difference between two audio files. The other mastering engineer in the article (Jason Ward) isn’t trying to “fool a null test” in any way. He does a different master than the original CD master for that particular song to try to counteract the unpleasant EQ artifacts that were caused when the original CD master was encoded by iTunes. Then afterward, because he did a really good job with corrective EQ, the null test showed very little difference (less difference) between his own custom “mastered for iTunes” version and the original CD master.
But the article does seem to indicate the intended point of your comment, that it does seem possible to counteract the negative EQ effect that normal AAC encoding has on a CD master. This doesn’t meant the null test from the video above was false. These were two different songs mastered by two different mastering engineers. When viewed together, the two tests seem to suggest that while it may be possible to correctively master for iTunes so that the end result sounds closer to the CD, the claim that it sounds closer to the CD in every case isn’t true.
