First, apologies for the fast-and-lousy reply -- I am on vacation,
and only have a somewhat touchy dial-up connection today.
Basically, I think GB & TF have a good idea, that more flexibility in
this area is desirable. But (in part perhaps because I use somewhat
different, although related, definitions of "accuracy" and
"precision" than TF does) my first reaction is that precision is
already represented in the normalized date. If we could come up with
a reasonable way to do it, though, an attribute for accuracy (once
called exact=, I think) would be a good idea.
On the other hand, my argument that precision= is not needed (which
swayed Council, BTW -- the precision= attribute no longer exists in
the development version of P5 on Sourceforge), does not hold true for
those who are not normalizing their dates.
Note: I don't think I can put up with using Sourceforge on this
connection, so I am only posting here for now.