Scientists often say an earthquake prediction should have location, magnitude, and time bounds. This is not because it is necessary to be useful, but rather because the prediction is then much easier to test, and more likely to be useful.

If one just said there will be an M6.3 earthquake this week, as struck L'Aquila, that's true almost every week! And would the occurrence of an earthquake corroborate the power of the prediction? Should the Red Cross pile up the supplies when a week is forecast to have an M6 somewhere?

It's odd that you philosophize about location, magnitude, and time bounds, as your predictions have none of these specifics well defined, and hence are very difficult to test for whether they work. Indeed, you have no idea whether your methods work, and certainly no proof to offer.

Thousands of earthquake predictions schemes have been proposed AND TESTED, and none work better than the ETAS model, with which you are probably not familiar.