Plans for improving Coverguess searchability?

Despite its accuracy limitations, the surfacing of CoverGuess in our catalogs is really cool. Are there any plans to improve its searchability, either within our catalog (so far as I can tell, we can find books matching a single tag but nothing more complex) or externally (so I can find all the books that have been tagged with a particular tag)? This would probably require a bit of normalization of the tags - at the very least making them case-insensitive so that "dog" and "Dog" both match the canine. This would seem to enable some of the more powerful use-cases for this data set - someone remembers a book where the cover had tentacles and a mushroom cloud in office cubicles? The data is there for it to surface The Atrocity Archives - but we can't search it right now.