Not long ago I was bitten by a problem with clash between exported names and method name resolution, which doesn't have a clean fix. Investigation of the problem showed that it could be fixed easily in perl source, and the patch I've concocted would probably even be a good candidate to 5.12. But I went further, and found that if I'll succeed in changing perl SV structure so that objects (as well as packages) will have ability to have methods declared (so called dynamic methods), not only the original problem would be solved elegantly, but also perl would get per-object methods.
I know also that many recent CPAN modules actually do just that, they provide a way to create per-object methods by re-blessing object on the fly to the dynamic package. So I'm unsure, I'd be interested in working on this problem, but I'd hate to end in a situation where the solution is unneeded. Partly because I've got no responses for this proposal on p5p.
So my question is about how new, possibly not-so-cool features in perl5 are needed, - has everyone gone off to perl6? Is trying to add new stuff in perl5 is flogging of a dead horse? I know that perl6 was born exactly because too many people wanted changes, but I can't say that I'm interested in perl6 as much as I am in perl5, especially given the status of the former right now. Possibly I've missed the train by five years, but what would be my options today?