I'm the designated ArtArt last call reviewer for this draft.

This is essentially a _pro forma_ update of the corresponding HTTP/2 document.  It's substantially ready to go.

Three nits:

1) The readability of the main paragraph of section 2 could, I think, be improved.

Instead of

   "Where [ORIGIN] indicates that the ORIGIN frame should be sent on Stream 0,
    this should be interpreted to mean the HTTP/3 control stream.
    The ORIGIN frame is sent from servers to clients on the server's control stream."

this should (unless I misunderstand it), be a new paragraph, something a bit more explicit, e.g.:

   Where [ORIGIN] indicates that the ORIGIN frame should be sent on Stream 0,
   this should be interpreted to mean the appropriate HTTP/3 control stream:
   that is, the ORIGIN frame is sent from servers to clients on the server's
   (unidirectional) control stream.

2) To match the HTTP/2 and IANA registry entries, the table in section 4 should read

    ORIGIN 	0x0c 	Section 2

3) Again in section 4, it took me rather a long time to chase down the actually relevant bit of the '[IANA] "HTTP/3 Frame Type" registry [HTTP3]', which is not in the referenced document at all.  Wouldn't a reference to the think itself, https://www.iana.org/assignments/http2-parameters/http2-parameters.xhtml#frame-type, be better here?

ht