From a business perspective -that here doesn't mean money but a widespread adoption of Drupal- I think the case for full multilingual support in Drupal doesn't really need much more explanation. For some interesting read, http://drupal.org/node/29645 [Mike Gifford: A case for better handling of multi-lingual objects in Drupal]

To have a good multilingual support doesn't mean to have some spare modules providing some nice features and flags. As the language in which you're browsing a site will affect every single part of the page, an important part of this functionality needs to be built in into the system, so all the objects can be translated to multiple languages without having to patch the core system.

This is very difficult to achieve -and takes too long- with small incremental patches. Thus, if we want a clean and powerful implementation, we need some work to be done at a time that addresses multiple small issues all across Drupal. The resulting 'patch' wouldn't need to be that big but, as it will impact so many core mechanisms, we need at least some general agreement on the view of a fully multilingual future for Drupal.