This is basic statistics:

If a correlation seems to appear, test it against more evidence not used in the original test. You have plenty of 7.9's, 7.8's, etc. Give it a try.

This is assuming a model most people don't buy:

"The higher magnitude earthquakes are somewhat easier to study because the triggering force has to be stronger to overcome the fault zone’s resistance to fracturing." Whatever gives you this idea? Especially when you are talking about a 7.8 vs an 8.1, it is not likely. And you don't seem the buy the much more obvious likelihood that tidal patterns should be mostly symmetric with respect to the near and far side of the Earth.

I have to mention that I've taken a first look at the 9000 largest events of the last 20 years, and see NO evidence of correlation of earthquakes with biweekly tides. So I'm not as inclined to allow 17 events to outline a pattern as I was last week.

John