A better measure, assuming that pennies facilitate value exchange[1], would be whether the cost to mint a penny exceeded the marginal increase in GDP[2] due to having that additional penny available.

[1]: This assumption may not be true; if they're worth so little that people lose track of them, they could actually make it harder to exchange value.

[2]: Making the GDP higher is also a very debatable measure, but I think this generalizes to other dollar-denominated measures of prosperity.