They are not impossible to track, but that would be relevant only if Mullvad were severely compromised — and even then, we would only be in almost impossible territory.
There are no central repositories as to the location of arbitrary banknote serial numbers.
Lets assume, for the sake of argument, that a cash-paying user were to make the mistake of paying every single time to renew the same suspicious Mullvad account using cash which was always newly withdrawn from cash machines from a banking institution which meticulously tracks them and is able to report from which location they originated (maybe even the card which withdrew them!).
In that case, if Mullvad were to be compromised (or if the targeted user was such an absolute threat to humankind that Mullvad were to agree to collaborate in his or her capture), it would only be possible if Mullvad's mail receivers were to either a) actively keep track of either banknote serial number and link it to a customer, or b) be fully aware of the requirement to make a note of it only of received to renew the target account.
Anything short of that and even the perfectly traceable banknote serial number just becomes one of hundreds? thousands? deposited by Mullvad in their bank accounts — assuming they don't even use some of them as petty cash if needed.
If a user of Mullvad were to reach that level of a threat model I would argue they would be much more likely to be caught by tracking of their sent mail, in the style of Ted Kaczyński.
> There are no central repositories as to the location of arbitrary banknote serial numbers.
Why do you say that?
All that's needed is banks tracking serial numbers and associated persons as cash leaves the bank and enters it. The serial numbers on American cash seem machine readable, and on each bill they are printed in two places near opposite corners - as if they are designed for automated reading.
It doesn't have to be perfect, logically infallible, alibi-proof evidence. You could build a pretty good graph of who is doing business with whom, especially by examining repetitions of the same edges. At worst, it seems useful for intelligence tasks and to obtain worthwhile leads to pursue.
There are no central repositories as to the location of arbitrary banknote serial numbers.
Lets assume, for the sake of argument, that a cash-paying user were to make the mistake of paying every single time to renew the same suspicious Mullvad account using cash which was always newly withdrawn from cash machines from a banking institution which meticulously tracks them and is able to report from which location they originated (maybe even the card which withdrew them!).
In that case, if Mullvad were to be compromised (or if the targeted user was such an absolute threat to humankind that Mullvad were to agree to collaborate in his or her capture), it would only be possible if Mullvad's mail receivers were to either a) actively keep track of either banknote serial number and link it to a customer, or b) be fully aware of the requirement to make a note of it only of received to renew the target account.
Anything short of that and even the perfectly traceable banknote serial number just becomes one of hundreds? thousands? deposited by Mullvad in their bank accounts — assuming they don't even use some of them as petty cash if needed.
If a user of Mullvad were to reach that level of a threat model I would argue they would be much more likely to be caught by tracking of their sent mail, in the style of Ted Kaczyński.