Zdenek Broz of the dicts.info project has done some interesting analysis of a recent en.wiktionary.org dump: http://wiki.webz.cz/wikt.png

This bar chart shows the number of translations within articles that have "Translations" group headings at all (note that most en.wiktionary.org pages don't have any). Accordingly, there are only 21031 translations to English, followed by 14341 to French and 13004 to German.

While en.wiktionary.org has nominally 313537 pages, the vast majority of these are tiny stubs with short definitions, but no translations.

For comparison, OmegaWiki (formerly WiktionaryZ), which is still pre-alpha software, currently has:

175990 expressions ("pages" in Wiktionary lingo) 11948 DefinedMeanings, i.e. concepts (comparable to the total number of separate definitions; Wiktionary has much more of these) 117 active languages 10964 English expressions 11277 German expressions 9342 French expressions etc.

This isn't an entirely fair comparison, since Wiktionary, due to its architecture, is split into many language databases. One could potentially derive more translations by parsing and combining data from the different editions. Of course, the lack of a machine readable format makes that difficult. Still, it can be said that Wiktionary, as of now, isn't particularly useful as a _translating_ dictionary. It's very useful as a _defining_ dictionary, though (arguably a very similar application to Wikipedia with much shorter texts).

On the other hand, the architecture of OmegaWiki seems to be validated by its steady growth. I'm fairly happy with our backend, but our frontend sucks; data entry is still much too cumbersome. This will be our big challenge for next year -- taking the project to a final, feature complete release that is user friendly. But I think we're on the right track.