Thanks -- that's very helpful and supports the position that the IETF is not now and never has been IANA. (Which in no way discounts the close relationship that the IETF has with IANA, and the historical origins of both.)

[…]

No one has ever said that IETF “is” the IANA. What has been said is that the IANA is a label for registries based on IETF standards. I strongly suggest that you read RFC 2434 here: https://tools.ietf.org/html/rfc2434 for a basic explanation of what IANA is and how it is related to IETF.

Next I would suggest that you read RFC 2860, the MoU between ICANN and IETF regarding the IANA functions for protocols, which commits ICANN to

“The IANA will assign and register Internet protocol parameters

only as directed by the criteria and procedures specified in RFCs…”

In RFC 2860, the IETF basically dictates to ICANN what the IANA will do to ensure that it is integrated with the IETF standards development process. The IANA is therefore an extension of IETF’s standards process, it coordinates unique entries in various registries created by the IETF standards.

If it helps, it might be easier to view IETF as an Standards Development Organization and "the IANA" (initially (in the early '70s) with Jon Postel as the sole "numbers czar" then later, more formalized and having multiple full-time staff) as its secretariat. That is, I believe, how "the IANA" has always functioned and my understanding (which is undoubtedly wrong) was that ICANN was created in order (among other things) to provide a legal entity for the IANA functions.