In 1975, headers were 40-50% of an average message. Now,
if they're 2% I'd be amazed (I haven't analyzed it in a while).
I agree with this (not necessarity the numbers) but the principle.
I disagree. Headers have exploded, and many messages (especially those
on mailinglists, which would be the bulk of all messages for many
people) only use text. Header-to-text ratio varies between 10/90% and
80/20%.
I will
say that mail quoting is a much larger part these days with the mail
bandwidth, especially the form of reply behavior where people reply at
the
top (no inline quoting, commenting) and simply keep addending the mail
thread.
This is evil. Too bad you can't shoot people who do this.
I think it would be good to address quoting and over-quoting in a new
protocol. If we can address over-quoting by embedding a pointer to the
original message rather than the original message, that should save a
good deal of bandwith and it makes it very simple to configure a user
agent to disregard this and not display it. For regular quoting this
gets a bit more complicated, but it's worth the effort as this way we
get to cryptographically protect the quoted text.