The following is purely hypothetical and should not be taken as advice or a suggestion.
Since all of the content wikipedia is free for public use as long as you credit it, what’s to prevent anyone from copying large chunks and importing them into Google Knols. Lets say you took the top 1000 pages from wikipedia, got some cheap labor from country with a weak economy like
india, china, russia, united states and have them copy/scrape all the pages and create knol pages for them. You could get around the dupe content issues by rolling back 3-6 months of revisions on wikipedia. Beg, borrow, spam, or rent yourself enough links to get them indexed. Wait 9-12 months for Knols domain authority to kick in and then turn on the adsense.
Again this is all purely hypothetical and not to be interpreted as a suggestion or advice.
Just saying I don’t think Google thought this knols thing through completely cause there are all sorts of holes in it, like mack truck and sherman tank sized holes, not teeny tiny mouse holes.