I often wonder how many 1918/7-D overdate
Buffalo nickels were minted. I know that there are no mintage figures on this and generally most other mint errors, but does anyone recollect any "intelligent guess" theories being ventured on how many there might be? I don't have my reference materials handy as I write this, but I think overall mintage figures for the 18-D was either 7.8 or 8.7 million, and the overdate is a subset of that figure. On a post from back in November 2011, forum member Danester posted this great photo of what looks to me to be an UNC overdate:

In his post, Danester pointed out that the coin was from a "later stage die" and that it had a lot of "die fatigue". The three questions that occur to me are:
Is die fatigue a kissin' cousin of "flow lines"?
How many nickels can one die generally knock out before it becomes too "fatigued" to continue?
How many dies were used that had the 8/7 in them?
It seems like if you knew the answers, or were able to make informed guesses, about those last two questions you might have a good feel for how many of these things might have been punched out.
The thing that got me thinking about this subject (again) was that I restored a dateless nickel last night and it turned out to be an overdate. (Yay!) I've stopped keeping track of how many of these I've uncovered, but it's something north of 30, I'm sure of it. That, in itself, tells me there had to be quite a few of these, after all, I'm just ONE guy. This particular nickel, however, was the first one I ever uncovered that had what I considered to be "flow lines" on it -- by the chin - (thus my flow line vs. die fatigue question, above) and that's what makes me think there might have actually been quite a few of these.
Thoughts?