Jest so Don doesn't have all the fun, I'll try to answer this one (at least a little bit).
Far-field earthquake triggering theories are "as old as the hills" as they say. It
seems one of those innate suspicions that humans have that if the ground is
shaking enough that instruments can feel it, then couldn't this result in some kind
of response (like an earthquake) wherever that motion occurs (which in the case
of large earthquakes is the whole world).
Back in the middle ages of seismic theory I recall the occasional discussion
of possible links between earthquakes. In 1886 after the great Greek earthquake
was followed within a few hours by a series of shocks in the Charleston S.C.
area and several days later by the large earthquake in Charleston, the local
papers were filled with speculation regarding a link between the two events.
Modern reinvigoration of the idea came from the observations that earthquakes
throughout the western U.S. were clearly triggered by the Landers, CA. earthquake
(Mw 7.6) of 28 June, 1992. This brought about a Science article co-authored
by a list of seismologists which nearly filled the first page (Hill was the first)
promoting the idea that these events were "triggered" by the Lander's earthquake.
There was a lot of discussion about this for about two years in the Journals,
until it was no longer considered a relavent topic. But, it did break the barrier
so that the USGS knee-jerk reaction to any question "Was there a relation between
earthquake X and earthquake Y" ...... "No way" became obvious for what it was
an unthinking response based on a lack of research and adherence to paradigms
which for good or bad, had permeated seismological thinking for generations.
About 8 years ago two notable seismologists Clarence Allen (California Institute
of Technology) and Frank Press suggested it might be possible to "predict" earthquakes
using far-field information. This paper which was published by the National Academy of
Sciences is a classic reference now. Since then many others have looked for
relationships between earthquakes and ways to use this information to help
determine the location and magnitude of the next one.
In general, there are not really any classical "originators" of this idea because
it seems ingrained in the idea of causality, certainly not like Wegener was the
originator of the Continental Drift idea anyway.
I am aware of many incarnations of using far-field theory to predict earthquakes,
including the far-field forecast I regularly post to this board. Exactly how Don chooses
his events, I am not sure. I think we need to accept his explanation
for now - at least it gives us something to hang our hats on.