I didn't take the time to analyze the author's methodology, but he takes the time to explain what he did and provides the Perl and PHP code he used for the benchmarks. It should be straightforward for someone to try to reproduce his results.