Thanks for the interpretation and ideas. I will work on that.
I have not "tested" the keywords because I really don't know how you can test something so simple. For example, if you sell german stamps for collectors, one would think that "german stamps" and/or "germany stamps" would be about as well targeted as one can get. I believe that my keywords are what those few people looking are using to look for it.
However, adding negatives could perhaps help.
I can certainly work on and test my ad copy. Frankly, I had not thought about that in terms of something that would affect ad position. (I really wish that they would simply rank on how much you are willing to pay.) I guess that because they are relying on their own concept of "quality", increasing click-through by improving ad copy would improve "quality" for them (and make more money for me).
However, there is a catch-22 in all of this. Unless and until I am on the first page, any testing/improvement work is likely to have almost no result. Thus in order to effectively test ad text, etc., I have to pay an average of three times what I used to pay to get onto the first page. Once on the first page where people can see the ad text tests, the testing will be more relevant.
Any and all ideas are needed and welcomed.