I agree... I think the test could be better if we simply selected which one we prefer.
Fun, nonetheless.
For those of you who want a better experience with this test, as I did, I imported the audio into Pro Tools and split it out to three tracks... A, B and C. They are all perfectly phase-aligned so you can take your time and really A/B(/C) them in a non-linear way like you ought to. Here is a link to download my session. I'll have it up for a few weeks.
http://dl.dropbox.com/u/6925090/HEAT%20Challenge.zip
Jason
P.S. Feel free to engage the trims and try some null testing. There are some interesting results.