Perl 6 is still 9 times slower than Perl 5 when parsing log files with a regex. I don't quite get why the feature which made Perl famous has now taken a back seat. Why should I take a 9x performance hit?
The Perl 5 core is hack upon hack, to the point that it's been mostly abandoned due to how hard is to change things without breaking anything. Perl 6 was born as a specification, which allows for a "first make it right, then make it fast" approach.
Unfortunately, the performance of regexes has not been improved much in the past years, that is true and on the radar. Yes, it can be improved (very much indeed).
Most other subsystems of Rakudo Perl 6 have been improved significantly, some by orders of magnitude. And then there's work on the basic runtime of MoarVM and its JIT that will benefit all programs running on MoarVM.
Perl with poorly-performing regexes is beta software, as far as I'm concerned. Why OOP, junctions and all the less relevant stuff has taken priority over regexes baffles me.