I must've missed an RFC that overrides prior RFCs regarding the To and From address headers. That "comment <address>" is commonly used doesn't alter the RFCs that define Internet Messaging. The To header is mandated: it must appear a minimum of 1 times, and a maximum of 1 times. The From header is mandated: it must appear a minimum of 1 times, and a maximum of 1 times. However, the comment token in those headers is optional: it may appear 0 or 1 times. Only the address token is required in those headers. Just because most e-mail users construct those headers to show "From: userA <userA@domain>" and "To: userB <userB@domain>" doesn't mean the comment token is required. It could be "We Cheat 'Em How <realtyspammer@domain>" . The comment token can be ANY string the sender wants to use, not their real name, not their name in the account, or anything identifying the person(s)." In fact, mailing lists often use a comment token that identifies the mailing list, not its admin, like "Garden Beauty Monthly Newsletter <noreply@domain>". There is no one named "Garden Beauty Monthly Newsletter" involved in sending out the mailing list. The only required token in the To or From headers is the address token.
Why would the recipient of a reply showing just their e-mail address not know they are identified by just an e-mail address? A spammer wouldn't know, because they puke out from temporary or bogus e-mail sources, but you should NEVER reply to a spammer to validate they hit an active and monitored e-mail account.
There is a problem if those headers do not enclose the address token within angle brackets. The address token is supposed to be angle bracketed; else, it become indistinguisable from the comment token. "To: user@domain" specifies only a comment token, and the address token is missing. Clients and servers might allow sloppy syntax (i.e., the address token is not clearly delineated with angle brackets), but sloppiness can result in delivery problems, especially with NDRs (Non-Delivery Reports) saying your message could not be delivered, because you didn't specify the address token. If SG is deleting the comment field, including only the address token, but not delineating the address token, then SG is generating invalidly syntaxed headers. Just because some clients or servers are sloppy isn't cause for SG to do the same.
https://datatracker.ietf.org/doc/html/r ... tion-3.6.2 from = "From:" mailbox-list CRLF
https://datatracker.ietf.org/doc/html/r ... ection-3.4 mailbox-list = (mailbox *("," mailbox)) / obs-mbox-list
and
mailbox = name-addr / addr-spec
and
name-addr = [display-name] angle-addr
angle-addr = [CFWS] "<" addr-spec ">" [CFWS] / obs-angle-addr
The address token is ALWAYS to be angle bracketed, not shown as a text string that could be confused as the comment token (aka display-name). If SG is changing:
To: Your Friend <friend@theirserver.com>
to:
To:
friend@theirserver.comthen SG is generating an invalidly syntaxed To header. There is no address spec in that To header, just a comment, so the message may not get delivered due to invalid syntax. To show in another syntax format, the headers should be:
To: [comment] <addrspec>
From: [comment] <addrspec>
where brackets indicate optional tokens (i.e., the comment token). The angle-addrs token is still required, and that means enclosing in angle brackets. Regardless of how humans might interpret a comment-less header with a non-delineated address spec, SG should still comply with RFC standards. Don't make the clients and servers guess which token is specified in a sloppily syntaxed header.