With regard to the three above mentioned, if, as Gouby suggests, the repairer forgot that the die would be in reverse order to the coin, then how come we don't get specimens where both the 6 over 8 and 8 over 6 are present on the same coin? Especially on the 33A and 39A.
If he forgot that the die would be in reverse order to the coin, then by logical definition, if there is an 8 over 6, there should be a "preceding" 6 over 8.
Gouby also thinks that the reverse order theory explains the 1862, 2 over 1 specimens, as opposed to the re-use of old dies.
OK, I may be talking a load of old cobblers, but I did mention that it was idle musings.