A first step of talking about the issues and how to do this (NSP performance)
seems like a good idea to me.
Mind, this was tried several times before with limited success.
We do, however, need updates to RFC 1291.
Vikas's efforts live on in GISD. Developing new and different ways of
measuring Internet performance is always a win, when said method tells
us and the consumer something useful. Developing a sales checklist for
ISPs is probably not the mission of the IETF.
We might also want to
look at the OP-STAT WG efforts that Henry Clark and Nevell have
been working on.
Note that what we're doing is *COMPLETELY* unrelated to 1291 or GISD. Seen
that hole and have successfully driven around it :-).