Given that people do hundreds of test encodes when they actually use things like x264, I think that if you want to say anything general about these encoders you have to do more than one comparison.