My website http://TheRawFoodSite.com fell from number 5 to 8 in the recent change. I was not hit as hard as others, but would like to offer a few observations to the Google engineers that I hope will improve rankings and thus search results for your users.
1) Content scraping is a major problem. Autoblogs and scrapers copy original material, but rather than link to it, they make it appear to be original to the scraper's site. Since such autoblog/scraper sites do this to content from many sites with no regard to reliability of the source. Thus they contain nuggets of valuable accurate content and garbage. This does a disservice to your users. Especially because these autoblog/scrapers generate vast amounts of generally poor content and frequently outrank the original source websites.
Here's an example, in 2003, I put up a website on raw food. Later that year wrote some pages on the Master Cleanse and put up the first Internet Bulletin Board on the Master Cleanse. I wrote a book about the Master Cleanse in 2004. The book has sold 110,000 copies and been translated into 8 languages. For 4 years, my website ranked number 1 for "Master Cleanse."
I just did a Google search on three 20 - 30 word quotations from my most highly ranked page. The first string (in quotations marks, so I'd just get the full, actual string) showed 599 pages. This was from a quote on the back jacket of my second edition and is also on my home page and the highly ranked page. This string actually includes "I thank Peter for ..." and has been copied exactly as is. The second quoted string is from another paragraph of this same dedication to my book. It is on 1020 pages. Then I searched a quoted string from a personal remark on my most highly ranked page and it is on 9 pages on the web.
If there were any way to mechanically track this copying, it would be a great help to your users as well as those of us who create valuable original content. It would increase our traffic and thus reward our work and encourage us to do more of it.
2) I've observed that many of the newer (last 3 years or so) sites have some of the worst grammar and much duplicative content rephrased to appear different, but just mentioning the same keyword phrase again without adding any new information. I know one autoblogger and saw some articles he paid to have written that were so poor I couldn't imagine having them on a website. Perhaps there could be a grammar check combined with a test for the percentage of duplicate content that would search out these sites and deindex them.
3) I notice Wikipedia ranks highly for my keyword, "Master Cleanse." Unfortunately, Wikipedia's editors' biases against alternative medicine make their article less than accurate. For instance, my book was mentioned as making the Master Cleanse popular again in the 1990's. My book was published in 2004. Even after I corrected it on Wikipedia, it was changed back to the previous inaccuracy. Last I looked, Stanley Burroughs, the developer of the Master Cleanse, was called an "alternative medicine performer." It's too bad you can't devalue their inaccuracies.
I applaud your efforts to try to provide relevant, valuable information to your users. However, I don't envy you the task. I know some of my observations are on the border between suggestions to validate data and policing intellectual property.
In case it's valuable for your analysis, my website contains a bulletin board with more than 20,000 members and 50,000 posts; a shopping cart; several ads for items that go along with the book; and no ad sense ads at all. There is no advertising for any other website.