You know, I started deriding knols because they had such a stupid (and to me, mildly offensive and potentially dangerous) tagline. I now find myself defending the effort, even though my POV is the same.
I've got to applaude Google for at least muddying waters that have been in danger of clearing, in terms of "authoratative" answers from wikipedia. What I like about the knols is that you can have the same thing defined multiple times, and so don't get the single answer that wikipedia appears to the average Web-Searcher.
The single answer is what most non-professional searchers are looking for, and the mirage of its existence scares me frankly. I don't have so high an opinion of most searchers that I think that they'll see the wikipedia entry come up in their search, click on it, read it, have the answer they need, and then go looking for more details that will make the answer less clear. Why would they? The same people know that a healthy diet will make them live longer, yet Hungry Jacks still exists!
I still think there are some fundamental flaws with the knol model, such as the option to earn revenue by placing ads against an entry. Also, it re-raises the question of how you can provide content as well as search content independently (Yahoo! Answers, anyone). Still, to me knols are trying to present an information landscape that is closer to reality, by making simple answers harder to get. Its not an intrinsically bad idea, any more than Yahoo! Answers or wikipedia are. Maybe the goal should be to force those 3 resources to be at the top of any search, regardless of the engine.