I would imagine that this is why thread form accuracy, minimum clearance at the pitch dia., true perpindicularity between the axis and face of the mating parts, and proper torque all work together to prevent, or at least minimize any shift after the assy. is made. If the flank angles of both the internal and external threads, are in fact parallel to each other, with the proper clearances at the root and crest areas, the proper torgue should, in theory load all of the threads equally. If they are both cut with the same lead screw that has a standard lead, and uses the same gear train, all of the threads should have an equal axial spacing, whether it is .06250",
.06253", or however it comes out. Minimum clearance at the pitch dia., and a good lubricant, perhaps a mix of teflon grease and fine grain moly powder, to prevent galling from friction.
There will indeed be an axial stretching in the unsupported length of any unthreaded areas. The cutting of a groove, either internally or externally will reduce the cross sectional area of these short areas, and concentrate
the tensile load created by the proper torque. Once that preload is made, the mating parts should reach a state of equalibrium.
Threading as close to the shoulder, with a threading tool that is ground to allow only minimum amount of flank length, on the shoulder side should minimize the length to be preloaded. If a clearance groove is needed, a fillet should be made as large as possible to spread the stresses created. Any rake that is ground on the threading tools, should be ground perpindicular to the cutting edge. This will allow a tool to cut rather than push the metal from the form. Grinding the rake perpindicular to the axis will narrow the effective thread form, which if this narrowing is not equal on both parts, will create a situation that will create line contact along the helix of the thread which may create issues of radial shift, or even an issue with harmonics.
About thirty years ago, I had a 16" engine lathe from the flat belt era. It had a taper threaded spindle, on the order of pipe threads, but about a 6 tpi thread. Now you are using an inclined plane to push two parts into a tapered hole, while loading the shoulder joint. For this you need either a taper attachment, or a CNC controlled lathe/turning center. Another factor to be considered is the finish of the thread flanks. A smoother finish will allow more torque to be applied to the assy. due to the reduced friction.
Reducing the effects of the harmonics, is from what I have read, mainly a matter of testing different barrel length and contour combinations. The days of knocking over a soup can with the old Red Rider are long gone! A salute to progress, and change for a reason, after unbiased thought and planing, the consultation of your friends(no beer here!) ;-)
Regards,
Bob