Rick Tomaska has written some about proof dies in the American series of Franklin halves but also with regards to Morgan dollars and says essentially that the "cameo effect" may be lost with as few as 10 strikes of the die. He has said very little changed from the late 19th Century. Another fellow with some firsthand knowledge is Daniel Carr, who managed to purchase a slightly older Denver Mint press and now uses it to strike medals, overstrikes, etc. in matte, ordinary currency and proof format. He is quite a nice fellow IMO and may be able to share some information, at least as it applies to somewhat later technology - although I would imagine much may apply to later 19th C. technology. I am not as excited by specific technology or microvarieties but do find in interesting that old dies would be pulled up and reused - the 1839s were an extreme example. I also wonder how accurate minting statistics, even with proofs might actually be. Where, for example did they come up with the figures for 1887 and 1893? Let alone, off years of record proofs!