I am not aware of any result of Hardy with applications to cryptography, but I'd be curious to be found wrong