RE to Alan Day: Yes the "hardness" of steel (measured according to Rockwell etc) as you say is a factor by itself and has also been an issue of marketing business - sometimes it seems more talk than substance.George Jones for instance advertised using extra hard steel while from subjective experience when filing reeds for tuning it rather seems to me the hardness was less than common Wheatstone reeds!According to Geoffrey Crabb the reeds made for Jeffries by his great grandfather probably were retempered between initial processing and final tuning with the attempt to achieve a more powerful sound. According to reports by players this seems to have succeded.The effects on *tone* itself may be something else and generally speaking the ability to articulate the tone is counteracted by procedures increasing the amplitude - so the end effect usually is that you have to choose, either a mellow/sensitive soft instrument or a rough but powerful one.

RE to Claire:A free reed itself produces a pattern with up to some 20 overtones,some "harmonics" but most of them not having harmonic/regular relation to the fundamental tone. What happens with the *tone* on the way to the ear from a free reed instrument mainly is result of filtering/absorption from materials and compartments.
Claire:"So type of wood, metal v wood ends, bellows material, pan geometry -- all have an effect".
Yes - this is an issue for speculation as long as not being properly investigated lab-wise but some experiments have been done so we better stick to those for the time being (as you say too)and there seems to be very little substance supporting importance of pan material. One important 'knowledge' is that experiments with accordins have concluded that contrary to much historic belief there is NO "resonance" of acoustic importance from the reed pan.

"Materials" of various kinds do have *effect* - hard to specify what- one curious detail: I have subjectively noticed that button type definitely seems to affect the tone character BUT one very important
issue is THIS: The subtle tonal qualities we perceive as *players* of the instrument are to the greates part a result from the *higher quantum of overtones* which are lost very soon by air absorption and the *tone* is perceived entirely differntly by the audience at 5+ meters away ! This is one reason that the difference between a metal end and a wooden end may be eliminated at audience location.It may be contrariy obvious and important when making recording when choosing microphone set-up.

RE SJ: Some of the major (avaiable) accordion research has been done and published by prof Gotthard Richter, Zwota, Germany