@Garulfo: Do you have a special use case in mind, or do you do this for fun?
"for fun" isn't quite right, although there's probably some truth to it. Actually, translating 6,300 keywords isn't fun, and it's even worst when you don't really master the material behind... The point is that I am sad to not be able to offer the use of ConTeXt (such a great piece of software) to friends and childrens. And ConTeXt complexity or ConTeXt documentation are not the first hurdles. When you start with a new software, a new methodology, having first to struggle with vocabulary in a foreign language is definitively a (BIG) thorn in your side (in fact, usually you just don't start). Many popular softwares offer multi-language early in their development (netscape / firefox, scratch, wordpress, krita...). ConTeXt did it, the infrastructure is ready, the translations exist, at least partially or at an early stage. And when I started to use it, I thought that it could easily be improved and updated without requiring any Tex / Lua skills. One remark : I was wrong, I spend hours to put all the content (from mult-def and context-en.xml) in a simple csv table and building quick&dirty scripts to put it back in mult-def.
With our small developer and user base we struggle enough with keeping documentation current.
Struggle likely appears about not-so-trivial needs. It's a shame not to use it in Italian / Russian / Chinese / French etc... for a letter, for a 10 to 20 pages report, for an internal note. All the more so as the software allows it. The point is that the interface just deserves a minimal update and documentation, and that, as you point it, the small developer and user base doesn't have time to do it.