Actually, I hadn't contested the completeness of the NOA down below M4. But those are some interesting stats you are presenting.

Out of curiosity, how much of the NOA catalog makes it over to the Bulletin of the Internation Seismological Centre (ISC). According to the USGS, they are claiming that about a third of the events on the PDE make over to the ISC (They just don't say what third).

See: http://wwwneic.cr.usgs.gov/neis/epic/code_catalog.html

You mentioned Reasenbergs. Is this the same person that work with Lucy Jones in developing the formulae for aftershock decay rates?

Lastly, your right in that analysis of a prediction doesn't need to be gone into in any real depth until that predictor starts showing a significant number of successful predictions that have some significance to them. Meaning that somebody who predicts an M3 will occur today somewhere in the world really hasn't shown anything because these happen every day somewhere in the world. On the other hand, if somebody could say that we'll have 20 M3's today somewhere in the world, and we do in fact have just 20 M3's for that day (and they could consistently do this), then that would be something to take into consideration.

Dennis