TorBrowser always reported as unique

[trinity-1686a]

TorBrowser messes up canvas fingerprinting in a way that makes amiunique says the fingerprint is unique, and different, every single time. Sadly it makes amiunique quiet useless with TorBrowser: there is no easy way to know whether the browser is truly unique, or just unique because of the messed up, ephemeral, canvas.

Would you accept a PR which reports Not supported when the canvas result is obviously wrong (like it contains way more or way fewer white pixels than expected) ?