I think there are a few things going on here.
The Talent Pool
Photo exchange is a fairly small site, so there are a limited number of people that will visit it. Of those, a reasonable number of them will predominantly either use the site as a reference resource, or ask questions. I’d imagine there is a minority actually answering questions. Getting people to answer their first question would seem to be the main initial hurdle (if somebody answers + gets a positive response they’re far more likely to answer again). Bounties may incentivise some people to post their first answer, but I think it’s unlikely as initially 10 reputation is a reasonable reward for an answer and the bounty isn’t guaranteed.
The Types of Bounty Offered
Looking at some of your bounties (such as the ones you’ve got running at the moment) you’re essentially asking for better answers for questions that already have answers (in some cases accepted ones). I think the bounty changes will help a bit here, in that at least you can state clearly what you are looking for from the bounty. However, it’s going to be harder to get people to contribute to questions that already have existing answers because they have to be convinced that they have something worthwhile to contribute and the likelihood of this reduces with each existing answer, so again this reduces the pool of people that are likely to offer new answers.
Amount of attention
It’s true, that higher rep active users may have answered questions without the bounty if their attention had been drawn to the question. Editing a post moves it to the front page, so to an extent there is already a way to get the attention of active users.
If you have a blog / know a community where other users can answer the question you can ask them to do so by providing a link to the question (this may even gain you an Announcer/Booster/Publicist badge).
Your badge idea may also contribute to higher rep users answering some bounty questions (although I would imagine this would be a fairly short lived advantage before those users collected the new badges).
Bounties aren’t guaranteed to produce a satisfactory answer (they’re a gamble). However, they should result in a higher likelihood of the question being view and hence answered (although as I’ve said to get a new answer people have to believe they can answer it). With that in mind, perhaps a better measure of the success of the bounty system is to look at the relative page views, once the bounty is added. This will show if the page is being viewed more + the bounty is working as an incentive.
Whilst the benefit of bounties on small sites may be less (questions aren’t rushed off the front page most of the time) I do believe they offer some benefit and have been awarded for some good answers, so removing them doesn’t seem worthwhile.