I used to contribute a fair bit to Wikipedia.
At some point, however, I became pretty discouraged by how often anonymous or inexperienced editors would alter something I had written in such a way as to make it objectively, factually incorrect. Often such errors are not obvious to the “casual” Wikipedian or those who police it for vandalism, so it falls to the area experts to deal with them. Unfortunately I do not have time for this. I don't mind people editing my writing, as long as they don't introduce inaccuracies.
Faced with the fact that I can't keep up with this article-damage, it's no longer really worthwhile for me to contribute content to Wikipedia. Why expend the effort needed to write something when it has a high probability of being ruined within two or three years?
After a lot of thought on this matter, I think that Wikipedia's greatest contribution to human knowledge in the long run is going to be as a “recommended reading list” on various topics. I encourage people to think of Wikipedia more as an evolution of the manually edited directories like the original Yahoo website or the more recent DMOZ Open Directory. However, I don't have much hope for the actual article content, except for material that falls into the realm of “general knowledge” – topics on which any high-school graduate could easily distinguish fact from error.
Therefore, I currently limit my participation in Wikipedia to adding links to external material. I encourage area experts to contribute external links to Wikipedia wherever possible. When relevant, external links seem to have a significant “half-life” even in the face of misguided editors. External link management is one area where the “Wikipedia social process” seems to work quite well.