Don,

Prof. Keilis-Borok predicts a high probability (50-100?, he says 80+ %) chance in 9 months, the Japanese forecast is 60% in 30 years. So KB is much more precise in time. But the Japanese are forecasting a particular event, while KB's forecast covers a wider region, so the Japanese prediction is more precise in space.

Parkfield had a prediction similar to the Japanese one, which failed to meet its initial expectations. It has proven difficult to find "characteristic" earthquakes, which are ones that recur at even intervals with similar areas of rupture each time, so predicting specific earthquakes is falling out of favor.

Progress has been slow, but if KB is right, and we can detect that events are coming in months rather than decades, it would be progress.

John