To me Perl was just Weird, to no particular end. Not the kind of mind expanding Haskell/Prolog/Lisp weirdness that opens up new possibilities. It just does roughly the same things as every other language, but everything is done slightly differently due to evolving out of the primordial soup of bourne shell and AWK filtered through Larry Wall's brain.
Perl and Python were similarly powerful and useful languages, but I could learn and start producing useful code in Python after reading an hour long tutorial. Perl took an order of magnitude longer, and remained more awkward to use just due to the Weirdness. There was a momentum building in the early 2000s toward competitors like Python and Ruby that were seen as less crufty and more modern.
Perl's developers seemed to agree, since they cooked up their own competitor to Perl, an entirely different language confusingly called Perl 6. The coexistence of Perl 5 and 6 made the Python 3 transition look like a cakewalk -- at least it would have save for Perl 6's almost entire failure to exist for over a decade after its inception. It produced lots of constantly churning specs and blog posts about register based virtual machines with native support for continuations or whatever, but no implementation of a language that anyone felt comfortable using for any real development. Meanwhile people kept using the ossifying Perl 5 for existing applications, and gradually transitioning away as they were replaced.
Also PHP overtook it for the "just FTP a script to $5 shared hosting and make a webapp" use case.
Wasn't Intel trying to do something similar in Itanium i.e. use software to translate code into VLIW instructions to exploit many parallel execution units? Only they wanted the C++ compiler to do it rather than a dynamic recompiler? At least some people in Intel thought that was a good idea.
I wonder if the x86 teams at Intel people were similarly baffled by that.
Maybe that's true for this project (implementing pointer chasing list interpreter directly in hardware), but it's much less clear to me why it would be true of the much more commonly remembered examples of "lisp machines" like symbolics, TI, etc.