I'm not entirely sure why this warning happens at the time that it does. Probably some heuristic that the implementor was't sure of.
The only files that contains the string "possible comment" are in the source medley/sources/CMLDEFFER and in the documentation file medley/docs/SEC3-IRMFEATURES-II.pdf (also in the medley repository) page 24
So you could set *REMOVE-INTERLISP-COMMENTS* to T and not get any warnings. (The warning was only useful when loading old code and most always ignorable).
Alternatively, you could 'convert' the comments to a newer Medley format -- In SEdit select the comment or the whole body of the definition and press alt-; (command ; on mac).
This form of comment takes the form (* ;; "a string") where the comment is represented in memory as a list of three elements, the symbol IL:*, a symbol in the IL package consisting of 1, 2 or 3 semicolons, and a string. Such forms are assumed to be intended as comments. This format is supported by SEdit so you usually will see the comment justified in 1 of 3 possible ways.
Because comments are stripped by the definer (the DEFFER in CMLDEFFER), you can put comments anywhere in code that is introduced by a definer.
DEFMACRO isn't really an "Interlisp macro". If you ask (EQ 'CL:DEFMACRO 'IL:DEFMACRO) you will get T, and if you ?? EQ to look on the history it will show as (EQ 'DEFMACRO 'DEFMACRO) from either kind of exec. The name of the macro has a package, but that doesn't determine the interpretation.
I found the stash of TEdit files that constituted the Medley 1 release notes; Ill submit a PR adding them.
Do we update the document, change the default warning?