Definitely something I'm thinking about. Maybe even some kind of "confidence interval" of confidence ratio could be added next to an estimate.