This is definitely an important subject and one that is not necessarily simple to resolve. There are many questions, such as where does Planetside's obligation begin and end, what is reasonable to expect of the user community, and what of commercially motivated authors and artists who may also write such materials on their own and for their own benefit?
For Planetside's part I can say we intend to provide a full set of documentation of Terragen 2. However what this entails is not necessarily what everyone wants or expects. We will cover, at a minimum, the full range of controls and settings, the overall UI, basic scene construction and a full node reference describing the specific functions and settings of each node. This does not necessarily cover how to accomplish any specific effect, an undertaking which would be extremely daunting if you were to attempt being really comprehensive. An advanced scene creation section may be included which could cover some of these things, but really by the very nature of TG2's flexibility and power it is impossible to document everything. Even one particular effect may be achieved in many ways most of the time, and there may not even be a "best" approach, depending on your specific needs. We will also be including a library of basic and perhaps some advanced effects and presets which should help, but again they would just be a starting point and a lot will depend on a user's ability to experiment and learn from that basis.
That being said I do think there is a lot of room for user-created tutorials on specific things, as BigBen mentioned. Further to that I think cyphyr's point that artists are the ones who can really make the most of these programs is very valid. I very much hope that we can generate enough interest in the user community and the graphics community as a whole to justify some books on TG2. As far as I know there is only 1 current Vue book and it is not even on advanced topics. I would like to see us do better with TG2, but of course it is not entirely up to us.
Calico I also think your idea of a basic structure for the documentation writing process is a good one, and is similar to how I have envisioned things in some ways. I am currently contemplating how best to approach the overall documentation as we get nearer to actually being able to write it based on a more final application (it would be wasteful to write too much based on the pre-release, of course). The simple truth is while I have a decent technical knowledge of TG 0.9 and TG2, my artistry, creativity, and general sense of the more complex interactions of things is not the best. There are definitely others who more consistently produce fantastic images and novel techniques. If that is any judge of knowledge or ability, and that may somehow be transmitted or taught, then it is worth capturing. I would definitely enjoy working with other talented people to bring the best possible documentation to our users, and I will investigate these possibilities as much as I can.
The other thing to keep in mind is quite frankly some people are simply predisposed to being good with systems like this, and with creating landscape scenes in general, others are not. This is not even a comment on the ease of use of Terragen or any other software - it is a simple fundamental artistic reality. An amateur but accomplished painter may still feel that the brush in their hand must not be the same brush, or there must be some obscure way of using it, that the master knows. But that is not necessarily so - they may simply see differently, or there is some other unknown factor in play. The simple fact is that Terragen is like any other artistic tool. Some can weild it better than others and no amount of schooling can make up for that. That doesn't mean one shouldn't try to learn or that lessons are not valuable, only that one should set expectations reasonably and not get *too* frustrated that some people seem to be able to do this naturally, nor expect those people to teach it, or even that it necessarily *is* teachable.
I think if there is one thing to learn, one practice that will further your knowledge and ability more than anything, it is the ability to experiment. Gaining intuition and always remembrring the freedom to try things without fear of the consequences is fundamental. A grounding in math can't hurt in playing with function networks too.

- Oshyan