This has significant implications for randomness and entropy.

I can already see it being useful for verifying entropy and for verifying anonymization of statistical data.