I suspended my experiment with the conversion optimiser today, as the editor doesn't get it and I'm doing bulk edits. Make sure you leave it going for at least a week before you draw any conclusions is all I can advise.
I use it only as a reference tool, taking some of their suggestions into account. I often find that most of that they suggest to optimize doesn't apply to our site. Furthermore, when working within a budget, some of their optimizations can prove to be useful, showing that maybe some reallocation is in order to get the most from your CPC campains.