Countries whose economies are under stress often debase their coinage - Nero in 58 AD, Johnson in 1965 AD, etc.
This is great as far as it goes but we don’t need 20 coin or 100 coin studies. In order to be statistically relevant we need thousands to be tested and compared.
I wonder if this technique could be used to identify fakes in major collections?
There’s actually a market for collectible ancient counterfeit coins.