Lein-droid author here. A few commenters asked about the performance and maturity of Clojure for Android, so I decided to answer in the top level.
The main problem of Clojure for Android right now is the large size of Clojure runtime. This leads to two issues:
a) long loading time (can take a few seconds even for the basic app),
b) applications often gets closed by the OS once switched to another, in order to free that memory.
Since ProGuard is no help in optimizing Clojure-compiled code, the issues should be addressed in some other way. One that is possible is to write source-shaking compiler that will strip the final APK of the unused functions. It's pretty hard to do with the current compiler, but if we eventually write Clojure-in-Clojure compiler (there is a proposal for this year's GSoC by the way), it may be feasible to add these optimizations to it.
The main problem of Clojure for Android right now is the large size of Clojure runtime. This leads to two issues: a) long loading time (can take a few seconds even for the basic app), b) applications often gets closed by the OS once switched to another, in order to free that memory.
Since ProGuard is no help in optimizing Clojure-compiled code, the issues should be addressed in some other way. One that is possible is to write source-shaking compiler that will strip the final APK of the unused functions. It's pretty hard to do with the current compiler, but if we eventually write Clojure-in-Clojure compiler (there is a proposal for this year's GSoC by the way), it may be feasible to add these optimizations to it.
And finally a shameless plug: if you are interested in trying Clojure for Android, here are the links to get you started: https://github.com/alexander-yakushev/lein-droid https://github.com/alexander-yakushev/neko