[ajross]

One possibility is that static analysis for supersonic waveforms is a whole lot easier if you have planar shock fronts (i.e. the shock waves coming off the airframe make simple polygonal shapes). Back when this was done with pencil and paper it was really the only option short of building and testing real objects, so if you wanted to do any sort of optimization you had to do it with airframes that produced big planar shocks.

But those big, single shock waves are, of course, perceived by the listener as a "boom".