Just a nitpick back: "sv_bless(SV* sv, HV* stash);" (perlguts); it really is the scalar that's blessed, rather than the other way around.

I think this is a point on which reasonable people can disagree.

I just recently read a discussion on (usenet or a mailinglist) about how the discussion of dynamically vs statically typed doesn't really make any sense unless you agree on the parameters before hand. I suspect this is sorta the same thing.