I work in a shop where a majority of the code is not what I would call production ready. Yet it's in production and has been for years; failing on edge cases regularly and failing on normal things mysteriously now and then. The variance of quality of code and systems that are in production in the various shops of the world is astounding and in almost every freelance gig I've had I've seen code that I would never personally allow into production. "Production ready" is a subjective concept which I've seen to mean "hobbled and dangerous" as often, or more often than "solid and safe."
Production ready for my business (publishing) is different than production ready for real-time derivatives trading. Production ready for my brother's business (software testing) is different from production ready for a non-profit. Production ready for a web development firm is different from production ready for an ISP. Production ready for a biology lab is different from production ready for an industrial automation shop. Production ready for a sporting good store is different from production ready for a university.
Within those industries, production ready also depends on availability and skill level of developers, maintainers, managers, and consultants, as well as their taste for risk, the amount of ancillary tools and materials, and considerations for existing code and systems.
None of these are binary, off-or-on, simple, single right answer for everyone everywhere questions.
I don't know. I don't care. I know what I need from Perl 6 to use it. I don't really care if it's ready or not for anyone else to use it, in the abstract.
If you need specific features or characteristics, you're welcome to file specific bug reports or make specific feature requests. As with any free software project developed by a community, they may or may not be accepted but you get the source code either way, and you're free to do with it almost anything you wish.
In the absence of specific questions and details, this degenerates into a debate over the semantics of abstract words like "done" and "exists" and "vaporware" and "delivery" and "dead" and "useful" and "usable", which is a lot of fun for epistemologists but is almost always an unproductive rathole for developers.
Edited to add: I get the impression that some people think that the intent is for Perl 6 to replace Perl 5 wholesale. I only speak for myself, but that's not my intent. I have plenty of code it doesn't make sense to migrate to Perl 6 right now, and it might never make sense to migrate that code. With that said, I think that Perl 6 is a better language as a whole, but Perl 5 has the better surrounding ecosystem right now. Yet being able to use Perl 5 modules through Blizkost is a tremendous advantage for Rakudo, and it changes my risk/reward calculation for using Perl 6 substantially.