JH> Here are two more than came through today.  I've turned off my feeds to
 JH> 2:221/6 & 3:770/1 for now and am just hooked up to 1:229/426:
 JH> 
 JH> https://paste.ee/p/hBhCR
 JH> 
 JH> https://paste.ee/p/nN9aV
 JH> 
 JH> Again, the kludges look the same (except for the path and seen-by lines)
 JH> but the message body has again been modified.  One message has a SLMR
 JH> line where the other one doesn't.

The short answer to your messages here is: These are not considered duplicate messages as the message content is different.  But I'll explain a little more detail for you if you're interested:

Mystic doesn't only rely on MSGID because some software or messages/networks may not use them at all.  Instead, Mystic takes all things into consideration in totality which includes: MSGID, actual message content, PATH addresses (to prevent importing circular distributed messages) and SEEN-BY (to prevent distributing circular exported packets).

As you noted these messages aren't really duplicates because the message content is different and Mystic knows this (despite what the MSGID alone says). Because it knows they aren't the same, it doesn't treat it as a duplicate.

I can understand why some may think a duplicate MSGID alone would/should be enough, but legacy software (especially in the art scene where Mystic was popular) didn't use them at all and other prominent software (like EleBBS I remember testing, among others) were prone to spitting out two completely different messages with the same MSGID in the same echo.

As a result of all of that Mystic looks at everything and will not block a message unless its very confident that it truly is a duplicate message.  In your case those messages are not duplicates in message content.

... Fer sell cheep: BBS spel chekker. Wurks grate.

--- Mystic BBS v1.12 A47 2021/09/23 (Windows/64)