I meant it was useless in the sense that we can't accurately determine the number of users from looking at on-chain data. As you've also explained in earlier comments, there is the issue of "multiple addresses - one user" and there is the issue of "single address - multiple users". If we ignore the latter issue, we can put an upper bound at ~1 billion users for Bitcoin due to the unique address count but that's just an upper bound, the actual number is lower but we can't know how much lower.