Bob and Peter,
Umm... at great peril I'll offer another point of view.
At the risk of simplifying your descriptions beyond what you intended,
I think you've defined two groups:
P: Concerned with privacy, most likely for personal or very small
group interactions.
R: Concerned with formal commitments and hence desirous of
non-repudiation.
In addition to these groups, I believe there is a large set of people
concerned with authentication but not with non-repudiation. If I'm in
a position of authority in some community, e.g. an executive in a
company, an official of a university, etc., I want to be able to send
a signed message which cannot be forged. The vast majority of such
messages are important when they're sent and lose importance as time
passes. Repudiation at some later time is an irrelevant issue; it
usually doesn't happen. And if does, it damages the credibility of
the speaker quite severely. Therefore, what's important to this group
is that there be some means of uttering something that cannot be
forged by anyone else. The minimum infrastructure needed for this is
that the community of listeners needs the public key of the speaker.
It's helpful to have some sort of identifier associated with the
public key, but that does not mean that the identifier has to fit into
any sort of civil or legal naming system, nor that it be understood by
any third party.
This does not contradict Bob's point that all of this can fit into the
same technical framework. However, if a significant portion of the
framework serves only one portion of the users, the other users may
gravitate toward a less elaborate framework.
Steve