Blocking JS is a fingerprint

IMO, that is true... but could potentially be misleading. It can still be better to block, even if blocking itself is a fingerprint. (edited here, fix bad wording)

It only makes you worse off if the bits of identifying info with javascript exceed the bits obtained by the single option of disabling JS. And that depends on how many ppl do it. It's the -log2 of the probability of that event. Let's say that only 1 in 100,000 disables JS. (It's gotta be more than that?) That is just under 17 bits of information from seeing that JS is disabled, b/c you're only unique to one part in 100K. So if the number of bits obtainable with JS enabled exceed 17, then it's a net win to disable JS!

I don't know what % of ppl disable JS. I feel it's more than 1:100K. That's only 1 person in a decent sized city. IDK what the right % is. But for any reasonable value I can imagine, it seems like a win. Far more bits of ID-ing info can be obtained through JS, than the -log2 of that %.

Also, the more ppl disable JS, the fewer bits are obtained by seeing that JS is disabled, b/c the log2(%-who-disables-JS) gets smaller.