Question: The reasonings of the UDP6 IP address field (BEP-15)

[tearfur]

Disclaimer: This question is purely inquisitive and not rhetorical at all.

Quoting BEP-15:

That means the IP address field in the request remains 32bits wide which makes this field not usable under IPv6 and thus should always be set to 0.

As I interpret it, it means that the IP address field should always be 0 when announcing via UDP6. This sparks a few questions from me:

1. Does it mean you should also match the address family of the HTTP &ip= paramter and the undelying packet? It is supposed to be the counterpart of UDP IP address field, right? But BEP-3 does not seem require matching address families.
2. It is technically possible to fit an IPv4 address into the IP address field when using UDP6, why not allow it?

In practice, I observed that a lot of clients only have 1 IP parameter setting for both IPv4/6, so matching address families is out of the window. I've seen people using the HTTP &ip= paramter to announce their IPv6 address via IPv4 and all other sorts of anarchic usage. All of these confuses me even more.

It'd be great if someone can give me a full picture about the purpose of this parameter and in-practice usage. Thanks.