Also, I'm not sure the fact that the rays that forms the sonic boom are curved makes the sound itself weaker, it just takes longer to propagate down the the ground. But the fact that the sound spreads out towards the sides of the plane should help, the circumference of the shock cone at the point where it touches ground is proportional to the height of the airplane, so I'd expect things to scale down because of that?
I found and skimmed this document, "Review of Sonic Boom Theory" by K.J. Plotking (http://adl.stanford.edu/aa210b/Lecture_Notes_files/AIAA-1989...), which has some helpful information.
First, it notes that the shape of the wave does change while it propagates, but it turns out that the effect is in the opposite direction from what you suggested. That is, close to the aircraft the pressure forms a smooth S (the "mid-field" shape), but then as it moves further away, nonlinear effects causes the wave to bunch up to create the discontinuous N-shape. (See fig 1 and 2 on page 30.) So it seems, in this respect flying higher is actually worse---most research about reducing sonic booms tries to shape the aircraft to be long and thin, in order to make the waveform smoother, so we'd want there to less time for the wave to bunch up again.
Second, the document actually has a diagram of how the over-pressure depends on altitude! In fig 10 on page 34 it plots the calculated overpressure from a hypothetical Mach 2.7 supersonic transport. As I eyeball it, the over-pressure goes from 212 pounds/ft^2 at 40,000 feet (airliner level) to 129 pounds/ft^2 at 80,000 feet (SR-71 level), so it went down by a factor of 0.6. (In a cute coincidence, that 60% is almost exactly what's predicted by the scaling law eq. 25 on page 16, even though that equation is only intended for hypersonic aircraft.)
I'm not sure what to make of Fig. 10. I'm sure they are accurately plotting their equations, but the impulse for the SST is a minimum at 50,000 feet and increases for higher altitudes altitudes. The impulse for the HST generally increases with altitude. OTOH they discuss the N-wave forming in the far field and maintaining its shape, but also broadening which seems like it would reduce the impulse just like the overpressure.
The references 31-33 have alot of data. Here is a link to the first one: https://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/196400...
PS - If you are getting three digits of accuracy eyeballing Fig. 10, you must have gotten a much cleaner copy than I did. ;)
But yeah, figure 17 on page 26 of the document you linked seems to be exactly what we are looking for, great find! One can see exactly how the waveform changes with altitude. Moving from 42,100 feet to 70,700 feet reduced Δp from 1.82 to 1.13, a factor of 0.71.
With regard to the SR-71 making record setting flights and flying at the limits of its flight envelope to do so(100,000+ feet or "the edge of space"), I think the sonic boom at that altitude may not be noticeable to human hearing.
I know the that the SR-71 and the Space Shuttle have produced very noticeable sonic booms, but what isn't answered is at what altitude did the sonic booms become noticeable or annoying. It seems to me that with all the things that reduce the the magnitude of a sonic boom that it may be possible to fly supersonic without being annoying and slow down to subsonic to takeoff and land. It might also be possible to design the aircraft body to promote dispersion of the N-wave.
Typo edit