On 10-8-96 Scott Mitchell and his father Larry contributed to the break-in debate by describing a process called "coaxing" wherein the fatigue life of a metal part is extended by cycling at lower stress levels. They referred to papers by Lehr and Herold in the German language.
The following paragraph on coaxing is from a textbook, Mechanical Metallurgy, Second Edition by George E. Dieter, Prof of Eng. and Dir. Processing Research Inst, Carnegie-Mellon U. in Pittsburgh. The emphasis is mine.
"IF a specimen is tested without failure for a LARGE NUMBER of cycles below the fatigue limit and the stress increased in SMALL INCREMENTS after allowing a LARGE NUMBER of cycles to occur at each stress level, it is found that the resulting fatigue limit may be as much as 50 percent greater than the initial fatigue limit. This procedure is known as COAXING. An extensive investigation of COAXING (G. M. Sinclair, American Society of Testing Material, Proceedings vol.52, pp 743-758, 1952) showed a direct correlation between a strong coaxing effect and the ability for the material to undergo strain aging. Thus, mild steel and ingot iron show a strong COAXING effect, while BRASS, aluminum alloys, and heat-treated low-alloy steels show little improvement in their properties from COAXING."
The cycling regimen to induce coaxing does not seem to be met by the break-in procedures described in recent posts. The reed material does not seem to be the type for which coaxing is effective. It would be interesting to dig up the Sinclair reference.