How does such a high level of privacy-o-meter score is achieved in every transaction made by Foxpup?
If anything, it shows Blockchair's Privacy-o-meter doesn't mean a thing! A weekly payment at the exact same time to the exact same (vanity) address shouldn't get a 95% privacy score.
I had to bump up the flux capacitor for this one, and I'm sorry for the grave-digging, but I just wanted to say that even after recently encountering (and reading) through Blockchair documentation regarding this "meter"[1], I'm still unsure how is this percentage calculated and why it doesn't apply more weight to this characteristic that you pointed out at the time. If we take Foxpup most recent timely transaction[2] the only thing that appears to be stopping from getting 100 % as a score is because they supposedly identified matched addresses (We identified which of the recipient addresses possibly belong to one or more senders.[3]).
If we compare a transaction details which had a privacy-o-meter of 0 %[4] vs. the most recent one from Foxpup[5], the only heuristics that it's being applied to the transaction from Foxpup is "rare_fingerprint" - which is just described in their documentations as "This transaction has quite unique technical characteristics" - while the other transaction has at least 4 heuristics identified (amongst other properties) that are deemed to be more easy to track for blockchair.
What I can conclude regarding this exercise is that, like you initially said, they are missing some sort of heuristic to determine if a particular recipient address has been receiving a fixed amount of BTC from time to time no matter what sender is being responsible for the transfer - such behavior could be easily linked/spotted by someone who ended up investigating the receiver address... I reckon that these kind of tools, while may look good for outsiders, may give a false feeling of security in some cases (like this one). I wonder if this heuristic would be easy to implement?