Nice resource, sinewave.
I've done some experiments with small print ads over the last few month, and while my results weren't that crazy, I found that roughly 80-90% of the ads were profitable. I had a virtual assistant calling a number of smaller papers to place the ads, which turned out to be cheaper in my case. There are a surprising number of Thrifty Nickel-type publications that will give you free or nearly free ads (like $2-3 placement) if you place the ads with them directly instead of using those nationwide ad placement services. My $10/hour assistant was able to place quite a few each hour once she got the routine down, so it was a good way to test it. To scale it, though, you'd almost definitely want to just use a placement service and create campaigns by region or type of publication.
Anyway, I made separate domains for each newspaper and I tested an offer with a payout around $5. Where I'm from, it seems like these kinds of papers are read primarily by old people and poor or lower middle class people, so I tried to test an offer that would appeal to that demographic - something that wouldn't be too crazy amongst the "$50 Goat" and estate sale ads.
The average placement generated $85 and cost $4. The run for most ads was 1 week.
As for the ones that didn't profit - I didn't get copies of the papers, but I almost suspect that the ads were never run. The difference in performance is just too striking, given the similarities in the papers.