Cryptography is a very old subject in History and while many people concentrate on the Computer Age cryptosystems, quite a few people can be interested in older cryptosystems like oldish ciphermachines (Enigma and Purple come to mind) or even ancien paper-and-pencil based ones (ADFGVX and others). Do we want discussion about these as well?
I think it would be tragic if this site were to focus only on Modern Crypto. There are so many interesting ideas in the older cryptosystems, and this is truly the best arena out there for such questions.
Besides, would it hurt the site? I think not - I think CryptSE will only benefit from encompassing all cryptography.
I think questions about older systems may be acceptable as long as they are "tell me about X" and not "let's talk about X".
To make this a strong site, we will have to ask and answer good questions about cryptography. As tempting as it may be to allow borderline questions, we'll have to remember that future visitors to the site will use existing open questions as justification for new ones. Accepting a weaker question because it expands the base knowledge of the site is an invitation to a stream of similarly weak questions.
I agree with Paŭlo that the Playfair question seems to be a good one, and if we can get more questions about older systems that are also good questions for an SE site, then I would vote them up as well.
I find the historic questions interesting and nominally relevant and I definitely would like them to be part of the site.
But I'm sure I'm not the only intermediate-level enthusiast professional who's a little bit annoyed with the undue depth of coverage which is often devoted to pre-digital cryptography, seemingly at the expense of the far more relevant modern topics. No doubt it's much easier to make this material approachable when you can relate it to non-programmers with a pencil and paper, so more of it gets written.
So let's just make sure the stuff of the past doesn't overwhelm the discussion to the point that it obscures the knowledge about current and future cryptography.