stevan_little
I think the issue you encounter is a combination of isolation (the more involved you are in the community, the more you are exposed to best-practices and idiomatic Perl) and the quality of some of the older Perl documentation and books. In some of these older documents there was an emphasis put on "cleverness through TIMTOWTDI" which really had a damaging effect on some people (read: the people whose code you have inherited).
Perl is an extremely flexible language, which in the right hands can be very powerful. However, as Uncle Ben said, with great power comes great responsibility. I personally would rather my language require a little more self discipline from me 90% of the time, so that in the remaining 10% of the time I can hack some really crazy shit to get the job done.
I would be interested to know what version of Moose you used when you made this judgement? Because, the recent 2.0200 version includes inlined type constraints, which provided a non-trivial boost in performance for the native types as well as custom types (provided you include an inlined version in your custom type). If you were profile your code again, you might find a sizable change (and if not, please feel free to send the profile output to the Moose mailing list for us to review).
There is a compile time penalty for Moose overall, but for instance, the Moose generated accessors typically benchmark very well in comparison. They will start to slow down when you add type constraints, but this is essentially extra code you are asking Moose to execute so you can't expect to get it for free.
I highly suspect you have been using MooseX::Declare or MooseX::Method::Signatures, both of which have a high performance penalty for method calls. These modules should not be looked at as being Moose, they are optional extensions.