To be meaningful, and to provide a means to uniformly measure and compare, it would indeed have to be measured at a particular spot with reference to the engine. Sound, like light, generally follows the "inverse square law", which says that the intensity varies inversely with the square of the distance. In other words, the sound intensity at twice the distance would be 1/4, the intensity at three times the distance would be 1/9, the intensity at 5 times the distance would be 1/25, and so forth. This would be without regard to intervening obstacles or things like a dense patch of trees that would soak up the sound in that direction. You would assume they would want a higher level of sound in the direction of travel, and since (I believe) non-Genesis engines generally can be used running in either direction as lead engine, I would expect a sound peak at each end of most engines (not equal to each other, probably, but at a higher level than off to the sides). A polar plot of an engine's horn SPL would be interesting to see.