Marc-Olivier Barre wrote:
OK, stay with me for a while and let's reason together.
Filtering based on List-Id is filtering based on something that the
mailing list server does. It relies on actions taken by the mailing list
server, which do not depend on what the email client of the sender does.
Therefore, it is more reliable because it does not change with every
different email client.
If I set the list to remove duplicates, I only receive the private
reply, which depends on the email client of the sender, which may or may
not do the right thing. Therefore, filtering is less reliable, because
it depends on all those email clients out there to do something.
By forcing a Reply-To header, all messages are forced to go through the
mailing list server, therefore something (such as setting the List-Id)
can be done to each message, which allows for reliable and simple
filtering: only one header, it never changes (unless someone changes the
settings on the mailing list server).
By de-centralizing the message routing, all that is lost.