It completely sucked to "down grade" the pages, but that's what I had to do to increase rankings for those pages based on the algo today, so that's what I did.
Interesting. I worked on the complete opposite direction. When I make changes, instead of thinking of the algo and what it might cause to the ranking, I improved pages just for the sake of improving them. The first time I did it (307 pages), the traffic went up by 12k visits per day. The second time (another 150 pages), too early to tell if there's any traffic change. Now, I am on the process of improving another 60 pages. Frankly, I hesitate every time I make changes. It's Risky. But what makes me doing improvement is the face of the site in front of the users (and potential buyers). I can't expect Google algo machine to rank everything I think is better.