Presumably that's testable (to a certain extent), even with only one event. Apparently there's something very specific about the profile of the seismic jolt. Jolts that have similar profiles undoubtedly happen around the world with some regularity, but there's something about it that agrees with what's expected from a strange nugget collision. (The article isn't very specific about that.)
The model of strange nugget formation should tell you how fast you typically expect them to be moving. That tells you roughly how far apart in time you expect the seismic jolts to be. (Call it an expectation value.)
Based on the frequency of single jolts, you can assign a probability to the chance of finding one event, over a given period of time, where two random jolts are closer together in time than your expectation value. If that probability is very small--the article fails to mention that probability--then you can at least point to the event as anomalous, and propose an explanation.
In order to call it evidence for strange nuggets, of course, more events are needed. Then you can measure the velocity distribution and see whether it agrees with the expected velocity distribution.