(For list moderators: my previous message was rejected as it was too long;
because of the quoted text below the message. So, just ignore it. Though,
you could definitely raise the limit to 1MB -- from 40kb.)
Sylvian, you've opened one very interesting and important question and it's
about the various states of the languages in the world.
Kichwa is far from being in the worst position, but it's also example for
the languages with which we will be dealing in the future.
I'd suggest the approach in the few phases, as well as it could be the
draft of the road for the similarly developed languages.
1) As you said, the first one has to be about local knowledge. It would be
good to list the categories about which the contributors would write. That
could be a common place for other languages all over the world.
2) In a year or so start writing the most basic scientific articles. I
think we should start with the primary school knowledge, maybe even move to
build the textbooks into the future Kichwa Wikibooks. After we complete it,
it would be possible for Kichwa children to be educated in their native
language.
3) Around that time we should approach Academy and talk with them about the
standardization of terminology. That would allow us to build knowledge on
high school level in five to ten years.
In other words, I'd tell you that you should go with your idea and start
collecting local knowledge in Kichwa.
The only other question is related to the MediaWiki interface. Is it
possible to translate the most common messages in it?
Based on your input I will start collecting recommendations on a Meta page.