Hello,
not much discussion here about statistics... well, it's not an exciting subject really.
I thought the point of blind testing was the sample size. One person is not enough, maybe you need more than 100. If more than, say, 60% of them opted for A you might be able to say that A is better than B, and if it's 50/50 then that's a pretty good indication that A and B are close to the same thing. I would venture that when only one person attempts AB you will not get conclusive results unless one of A or B is truly terrible.
It's hard to get 100 audiophiles into the same room, and you would need good security to prevent fights breaking out.