An anecdote:
Back in the 70s I was charged with manufacturing and calibration of microvolemtric lab equipment to the 6th place (microgram by mass). We decided to mold the bodies of these handheld piupetters and dilutors out of PFVD, a remarkably nice, hard, moldable and machinable fluorocarbon with full chemical (except fluorines) inertness. After being in the market across hundreds of labs internationally we found that a percentage of the bodies developed a small but VERY conspicuous crack that, although innocuous, was not acceptable. We found that indeed the PVDF, like its softer cousins, cold-flowed across time. In that sense one could imagine the "rearrangement" of molecules mentioned and derided above. We found that we could accelerate the proicess by simply applying heat and pressure for a few minutes (as in a clinical lab/dentist's autoclave), wherein the PVDF then quickly reset due to the thermal shock...and then remained completely stable for many years (as far as could be ascertained). Is this cold-flow "curing" part of the so-called burn-in (npi) audio process. I doubt it. But still...?
The other aspect I wish to mention is the possibility that break-in is associated with the build-up of statis micro-voltages when handling cables in dry winter rooms. These charges have to quench to ground. My latest experience with my (ArgentPur) cables is that speaker cables may require 0-100hrs burn to fully "open up", whereas folks are saying that ICs do not. This leads me to question whether the delta is correlated with mass or length (spkr cables are bigger/longer), or usually on a static-charged rug!
YOUR thoughts? Thanks. ErnieM