When I went to see the state of perl a few months back I noticed the language is maturing, CPAN is still strong.
I think they have a PR problem.
The real issue is that a new programming language can be quite an investment. Some people learn lots of languages but many people actually stick with just one. Perl has nothing compelling over the other languages. You don't really lose anything buy picking e.g. Python or Ruby instead. So people tend to pick a language based on what cool new thing is written in it and that cool new thing is never in perl. Perl always rushes in later with a "me too" copy but it's never where a new breakthrough happens.
I think the GP was actually trying to say that the traction is missing for an otherwise nice and mature/stable infrastructure. A couple of "shiny" projects that boldly host with a powered by perl on their page might bring in fresh blood..
As a community, we tend to like to really think about things before implementing them - which means that while new ideas might turn up in perl a bit later than in other places, they tend to turn up in a well thought out way that we can confidently maintain backwards compatibility with over time.
When you're writing code that may still be in production in five or ten years, this tends to be an advantage.
I'm a perl5 hacker at heart and intend to continue to be.
The latest version of perl5 is version 12.3, released rather recently.
The websites for both languages now make this clear, but we'll probably need to rename them both to "-something- perl" for people to stop being confused by the names. Maybe in the next year or so.
As someone who worked at a shop that migrated from python/Zope to java/hibernate/GWT during the "zope 3 WTF?" days, I can attest that the confusion created by poor naming choices, and the lack of clarity in product roadmaps, was indeed a factor in deciding to migrate entirely away from Zope.
The first is that Perl 5 (the mature language) suffered greatly from the premature announcement of the Perl 6 project, more than ten years ago, which still hasn't shipped anything as stable and mature as, say, Python 3. (Several years ago, someone at my company showed up to work carrying a book entitled "Perl 6 Today". I asked where he'd gotten the time machine.)
The second is that if you look at blog posts and the like from inside the Perl community, such as it now is, you get the impression that it's the kind of crowd that thinks they're smarter than everyone else because no one else gets their inside jokes. Which comes from the top --- Larry Wall seems to think he's really clever when he justifies some of Perl's more dubious features by pointing out that natural languages do that too. But natural language has a full human intelligence at the other end sorting out the ambiguities, and even then, we still sometimes get it wrong. "Rakudo Star" is not that smart, much less Perl 5.