The Wikipedia article about mutation testing mentions an interesting insight about mutation testing. One of the goals of the analysis is to identify weak tests (tests that never kill any mutants). It could be a nice addition to the tool to keep track of the best "return on investment" tests and the worst ones.