Benchmarking different versions of Perl (or rather the same script written in different versions of Perl) is easy: just install all these Perl-versions and run your script in each one and time its performance. It would make for a nice statistic, but its value would be next to useless. All it would show you was how fast a script ran under various versions of Perl. It would tell you nothing about Perl as a whole.

I'm quite sure that you can write a Perl 4 script that outperforms a Perl 5.8 script. Should we therefore all turn back to Perl 4?