I don't think we're talking about an interlegibility criterion here, at
least that's not how I understand Jimbo's proposal. Certainly nobody objects to separate wikis for the Scandinavian languages. The issue is not whether native speakers of one language can read material in another language. It's whether there are native speakers who actually read and write in this as a literary language at all, as opposed to reading and writing in some "other" language of which theirs is merely a dialect.
The criterion of Jimbo is not incorrect. Some philology use the same criterion to distinguish tongues between dialects.
Expanding on Jimbo's point, here's the problem with dialect Wikipedias (leaving aside, for a moment, the thorny problem of deciding whether something is a dialect or a language). Dialects tend to be in a similar
position to constructed languages, in that they have no pre-existing literature and their orthography has to be invented to a significant degree. When the standards these establish are missing, we really have no factual basis from which to write in such a language. Inventing spelling or grammar while you write the encyclopedia, just like inventing facts while you write the encyclopedia, is perpetrating an intellectual fraud and a hoax.
This is not correct: same tongues can became dialects. The sicilian for example had a very important literature with Frederick II, Duke of Swabia. Also the occitanic at moment is not used in literature, or in law courts, or in officials circumstances.
It's wright that a dialect became tongue where it enriches with "registers", but a tongue can become dialects when it loses them.
Ilario