1) The top 1% of programmers will indeed go to the tools that make them the most productive, which are not necessarily the tools that happen to be currently most fashionable with the youngsters. Claims that Ruby is more productive than Perl are vaporware. Ruby did have an advantage for a while, but now Catalyst, RoR, and Django have now very similar feature sets (only you get CPAN as a bonus with Perl).
2) A language that is currently gaining momentum is not necessarily a language that will still have that momentum in 5, 10, or 20 years time. Perl popularity reached its peak years ago, then slowed down and has now settled and remained constant for a number of years. Given the kinds of mission critical applications that have been written and are still being written in Perl, with all my respect for Ruby (which I think is a fantastic language) I think it is a safe bet that in 20 years' time there will still be more Perl programmers than there will be Ruby programmers.