Of course to reverse the trend Perl needs more programmers etc., but Perl for its own sake isn't really a compelling argument. The question really isn't "What's required to make Perl a more common language?" The question is, "Why should anyone choose Perl over another option?"

Certainly there's a "best tool for the job" approach that might suggest Perl for certain applications, but most of the time developers are saddled with using a single language for a task. Like choosing a hammer: You'd like to choose the one best suited for a job, but that job requires driving both 20d framing nails and brads and everything in between. Carrying an 18oz for the framing, a 10oz for the brads etc. isn't practical, so you choose the 18oz because it's the only one that will work for framing and you can still use the 18oz for the brads, you just need to use a light touch and be more careful.

All that said, I take it this article may not be entirely serious. If it had not been written by Larry Wall, I might take it as complete satire, as examples of the exact wrong reasons to champion a language.