Getting started (going in circles)

Hi all,
 
I'm wondering what it takes to get started with Wyona.  I have to admit to having tried to get it going in a half-hearted attempt about five times now, and, well, I'm still without an operable instance on my machine.  Sure, it's booted, but now what do I do?!?  I've seen the Xopus demo, but that's running on their servers.  I want something that I can put a JPDA debugger against and look at code!
 
The thing that keeps bringing me back to looking at Wyona is the integration with Cocoon.  And I am starting to like Xopus more than I like chocolate milk.  But these (with the exception of Cocoon and chocolate milk) seem to be as elusive as the woman I would like to meet.
 
What's more strange to me is I used to have a better understanding of Wyona than I do now.  If the docs have been growing, they have been getting less organized.  For instance, there are no documents called "getting started", "quickstart", etc. that are available from the Wyona home page.  There are no architectural documents with goofy block diagrams or UML.  If I wanted to apply Wyona to my own project, I would have to understand it through brute-force osmosis of what it is doing on the sample sites.  
 
CMS is not rocket science, but it's not really very efficient for all of us to expend all this energy in parallel to evaluate it when we could be using that energy to extend it and make it better.
 
What I have gathered is Wyona is a very simple concept that uses a very small amount of code to pull off really cool stuff.  I was reading an article by Stefano about the early days of Cocoon, it sounds similar.  What I ask is whether the success path of Cocoon could be emulated.  I believe some of this has to do with documenting a reason that people will want to contribute to the project.  Xopus makes things simply unbelievable.  That it is so hard to get ahold of really breaks my heart.
 
Documentation is not fun to generate.  I don't do it either.  But I am not trying to get momentum around an open source project, so I don't have to worry about it.  Wyona is, and I think the lack of elaboration on the vision and the architecture is really holding it back.  Given the demand for a robust XML CMS and the success of projects like MMBase, I can only imagine that the lack of shared vision is what is messing up our plans.  
 
Could Wyona.org be published with Wyona?  What will it take to bootstrap ourselves?  I see the power power of a CMS so content can be contributed by anonymous authors and approved by gatekeepers (editors).  With Xopus running, it would be a walk in the park to edit content.  We should have the best documentation of any site going since everyone becomes a contributor, right?
 
Stefano and company met the other day to talk about CMS as well, I wonder what they came up with?
 
Sorry if this has very low signal-to-noise...
 
Brian