Hmm.. when you think about it, randomized may be the way to go simply because the other goals are virtually impossible to achieve.

OTOH, randomized with certain bounds is also a problem if it can be statistically guestimated which bounds were applicable.

nostr:nprofile1qqsxugjjylrmak4hvhn5k2ddnqgvsereprgq9sdeunp3jhxa58dgafcppamhxue69uhkgctdw4eju6t09uq3yamnwvaz7tmnv968xct8v5h8s7t69udlz000

I don't trust any of these claims.

That's really cool. It would be even cooler if you posted a link for us.

Love brave. Love EFF.

Link for those who want to test

https://www.eff.org/pages/cover-your-tracks

Pixel phone, standard android OS, brave browser got the same results as you