I normally run eBay/CJ, a 300x250 editor kit ad in the center of my photo gallery pages on one of my sites. CJ reports a little over 100 clicks every day. No impressions are reported but my stats show about 6,000 impressions per day for those pages. Yesterday I did a little experiment. I swapped in AuctionAds in place of the editor kit ad, a 250x250 ad block which contains 2 ads and looks quite a bit nicer than the editor kit ad, better blending and formatting. I got 23 clicks on 6,600 impressions. The editor kit and AA ad blocks are virtually identical, same keywords, same placement, although AA is somewhat better looking. But AA gets 1/4 the clicks of the editor kit ads. So why is that? Still bad reporting like we saw a couple months ago? Perhaps not. I think it's the caching. When I spot check some of the pages I often see totally irrelevant ads that have nothing to do with my keywords for that page. That was always an issue in my mind, the caching. I think that's why editor kit performs so much better for me. I can use long tail keywords, some of which are quite lucrative, and relevant ads are served. Not so with AA. The eBay editor kit performs much better for me than AuctionAds. Your mileage may vary.
goemark, Speaking of keywords and such one thing that I'd really like to see AA add to their program is the ability to restrict your keywords to a certain ebay category or categories. Another program I use called phpBay which uses a CJ ID does this and I love it. I still manage to make some good money from AA, but it would be better if this was available... would also help to limit irrelevant ads as well.
Changing your ads can get them noticed by regular visitors, as change is often a cure for ad-blindness. Maybe leave them for a month, then swap back and forth every couple of weeks. Though, don't just randomly rotate your colors. If your users constantly see a new color/format on each visit, they will soon become blind to the change. So manually do the altering, and only do it once every few weeks (or even months).