However, my opinion is that's a long way off since there is a lot of work being done, especially with CPAN to extend what you can do with Perl.

Actually, that's part of the problem with Perl's waning comparative popularity: languages like Perl, Python, and Ruby gain a lot of mindshare traction via their suitability for the Web, but CPAN actually acts as a demotivator in that realm. When people come up with new code in Perl that they want to share with the world, they stuff it in CPAN — but all those people out there looking to use Perl on their shared-server webhosting accounts are often pretty badly disenfranchised by that approach because they can't just go around installing modules willy-nilly when they often don't even have shell access, let alone root access, on the servers.

That doesn't even bring up the problem with dependency hell that CPAN-wrangling can often impose on our otherwise apparently simple solutions to commonplace problems. CPAN is great, but it suffers some issues, and these issues most affect the most fertile ground for growing Perlists.