I've been told by a Google engineer (on G+) that IP addresses aren't used for identity purposes. Whether or not they're aggregated at other levels ... would be interesting. I'm not sure whether or not browser fingerprinting is used at all, though I've suggested some reasonably non-invasive ways in which it might be applied for abuse and harassment mitigation, say: hashing fingerprints and storing them with the complainant's account in order to detect sockpuppet attacks.
You couldn't reverse the hash to determine the fingerprint, but if fingerprints matched, you'd have identical hashes. Splitting out different data records might help create probabalistic matches (e.g., n of m hashes match for an identity probability of p).
Tor would of course address IP issues. Privoxy could handle browser fingerprinting.
You couldn't reverse the hash to determine the fingerprint, but if fingerprints matched, you'd have identical hashes. Splitting out different data records might help create probabalistic matches (e.g., n of m hashes match for an identity probability of p).
Tor would of course address IP issues. Privoxy could handle browser fingerprinting.