> passively learning about each of the devices on your network and tracking them over time that it's foiled the plans of many site operators and nosy organizations looking to do just that.
Again passive learning (i.e. your device just existing and getting scanned) never happens in either scenario as inbound scanning is not possible in home or office in either IPv4 setups or IPv6 and it has nothing to do with where NAT is at play inbound initiated sessions aren't allowed in either case. I.e. just because you send a packet to <IP> does not mean a home or corporate router is going to allow it in, it's going to check if that conversation exists and if not it's going to check if the conversation started from the inside or the outside. If the latter it gets dropped, routability be damned, as it'd be insecure to allow anyone to connect to anything internal just because they sent a packet to that IP. This tracking is done at the L4 level, i.e. just because you opened a TCP session to a server using some high range outbound port doesn't mean that server will be allowed to send a packet back to you on e.g. 22 SSH it only means that specific tuple (ip:port:ip:port) is allowed bidirectionally until the session is closed by either side or times out from inactivity.
I'm not exactly sure which ways IPv6 makes it worse, as mentioned pretty much all devices (Windows, Linux, Mac, Android, iOS) use temporary IPs and have for well over a decade. Nobody is tracking individual user clients by IP address alone, be it v4 or v6, they are doing it with the troves of data in the higher layer protocol information where they can identify each device uniquely.
You can dislike IPv6 if you like but you shouldn't just make fear mongering claims of why it's bad for privacy if you don't have enough experience with it to know how it is designed to avoid these very issues.
Again passive learning (i.e. your device just existing and getting scanned) never happens in either scenario as inbound scanning is not possible in home or office in either IPv4 setups or IPv6 and it has nothing to do with where NAT is at play inbound initiated sessions aren't allowed in either case. I.e. just because you send a packet to <IP> does not mean a home or corporate router is going to allow it in, it's going to check if that conversation exists and if not it's going to check if the conversation started from the inside or the outside. If the latter it gets dropped, routability be damned, as it'd be insecure to allow anyone to connect to anything internal just because they sent a packet to that IP. This tracking is done at the L4 level, i.e. just because you opened a TCP session to a server using some high range outbound port doesn't mean that server will be allowed to send a packet back to you on e.g. 22 SSH it only means that specific tuple (ip:port:ip:port) is allowed bidirectionally until the session is closed by either side or times out from inactivity.
I'm not exactly sure which ways IPv6 makes it worse, as mentioned pretty much all devices (Windows, Linux, Mac, Android, iOS) use temporary IPs and have for well over a decade. Nobody is tracking individual user clients by IP address alone, be it v4 or v6, they are doing it with the troves of data in the higher layer protocol information where they can identify each device uniquely.
You can dislike IPv6 if you like but you shouldn't just make fear mongering claims of why it's bad for privacy if you don't have enough experience with it to know how it is designed to avoid these very issues.