@helmut_meukel
In the context of this statement โ one shot! โ the speed of sound is consistent during the duration of the shot.
Unless...
-the shot is made over a sunlit golf course, where the rising plumes of warmed air over the bunkers gives small variation of density, thus speed-of-sound
-the shot is made over an area of mixed dry and damp ground, where the rising water vapour has the same effect (yes, water vapour is less dense than air)
-the shot is made with significant elevation change
-the shot is made through an area partially shaded from the wind by buildings, as this gives variations in the SoS between the moving and calmer air
-the shot is made through the thermal plume from a fire, chimney etc (where not only the temperature and thus the density, but also the molecular composition of the air with more CO2 has some albeit small effect)
Indeed, there are many possible causes for density changes in the air which, because of the dependency invalidate the 'without change' clause so, strictly, joyR is correct. However the changes are all small and generally negligible, and this whole argument could have been avoided with a little more careful wording, like '... the speed of sound is consistent without significant change.
On the gripping hand, peri-sonic flight will be subject to greater buffeting leading to accuracy loss, so you would want to avoid a ballistic combination giving supersonic flight at barrel exit and subsonic flight at the target.
Anyway, how fast does a bullet slow down?
And, the thing that struck me first on reading is a little datum I remember from years ago, that in wood the speed of sound is different with the grain and across the grain.
After all that, I'm just going to have to re-read John Ringo's Unto the Breach again, for Lasko's long, long shot - and that means re-reading the whole series.