For that matter, does attenuation actually slow sound down? The speed of sound at a given temperature, pressure and humidity in the Earth's atmosphere is a constant, isn't it?

I'd tend rather to believe that the "shortest distance" between two points for a sound wave in the atmosphere isn't the same geodesic you have on the Earth's surface.

My suspicion would fall on the jet stream or the Coriolis effect which mostly but not entirely causes it. The sound wave was travelling considerably more "east-west" than "north-south", which certainly makes it possible.