Works very well, especially in v. 1.1.1, which adds many interface enhancements as well as fixing a couple of crash problems.
I don't know why anyone would have problems understanding how to use the application; it seemed quite simple to me, but some usage enhancements would be nice. Fortunately, the developer listens to suggestions--about half of the ones I've made have already been added (and I've gone a little overboard with suggestions).
Perhaps the main reason why duplicates might not be found is that, AFAIK, the application only matches duplicates of equal size. It'll find a single image multiply saved in in JPG, TIFF, and other formats, but won't match them if one is 640x480 and another is 1024x768. So far, that is the single largest drawback to an otherwise very good program (and one that is getting better).
I have a LARGE collection of photos that is rather disorganized, including sets that I cannot now identify, but might be duplicates of ones that were previously labeled and archived. However, since we're talking 89 *thousand* photos... well, just try remembering which ones! I've been able to cull out thousands of duplicates.
To explain my rating:
For "Features" to go to 5, it needs
a) the ability to match images of different sizes, and
b) the ability to rescan a particular result, adding new images. Now, you can add a new, separate directory, but can't redo already scanned directories that had new content added (matching only the new images against the existing database).
If the little interface suggestions I've made are added--e.g., when one of a pair of duplicates is deleted, for the next pair to automatically be selected even if in different folders--"Ease of Use" goes to 5.
Right now, it's helped do things I couldn't practically do before.