Believe it or not, I actually don't go for DNs being used as the
primary identification means for internet community UA products.
But you can play and decide yourself, if you are an academic and
non-commercial researcher:
In the UK portion of the PASSWORD security pilot, we developed and
experimented with both approaches (DNs at the messaging UI, versus
addresses) with a graphical PEM over MH/SMTP, PEM within a X.400
security UA, and also other security protocols linked with BBN's Slate
UI and Message Store bound to a secure organizational X.400 Submission
and Delivery Entity. Colin Robbins' company productized a version of
the JANET X.400 UA showing many of the DN versus address
functionalities. Its a very friendly, and very powerful demonstration.
The security demonstrator version is available in an internet binary
pilot form from him for researchers, though the pilot support period
may have ended. Im led to believe you can substistute the UCL PEM
filter and Internet security material for the GMD equivalents, so you
can play and pilot twice for just one (non-)payment!