Throwdown58 I agree in a sense. It is surprising how much knowledge about weapons people had in the later half of 19th century. The problem is that this knowledge was scattered through various books, magazines, newspapers etc. There was also a lot of false myths people believed in too unfortunately. 90% was forgotten then rediscovered and given new names. I don't know when barrel harmonics was defined as a term and studied systematically , but I have no doubt people back then knew barrels bend in various elastic ways during firing. One could say their experiment and your explanation are both different descriptions of the same process.
There is however one thing I don't fully agree with. One is obviously the "time wasted" on resoldering look at my other thread and you'll see I'm on the 4th resoldering of my set of double barrels now
Second is the ability to replicate. Today there is pretty good knowledge of soldering. One can imagine design of a repeatable process that reliably joins two barrels on full length is possible (using silver solder for example like Pedersoli). Then if one knows the correct principle and all the variables like pressure curve and time the projectile spends in the barrel it is possible to simulate stretching of steel using numeric methods (FEM is used to simulate loads in architecture, including vibration analysis in machinery etc). Then manual tweaking would probably be required to just get that final bit of accuracy that is lost due to tolerances in manufacturing.
The reason why it is not done is that the market for double rifles is pretty small today. The investment in writing the right design software would run close to a million $ probably(I'm a software developer by trade) . When you have customers willing to pay $50k per rifle, but you sell only few per year it makes sense to spend some time soldering /desoldering rather than investing in software.