I think that is too harsh. I often put scalar in when it is not neededfor the same reason that I use parens which are not needed - I know that the code is likely to be edited by someone who has not memorized the various contexts.

Now in a perfect all Perl world, I would agree with you. But when my code needs to be understood by programmers who are not primarily Perl programmers, well I am confident that I can teach them to remember that Perl has both scalar and list contexts, and they do different things. But when they have to edit my Perl after a month in other languages, I do not believe that they will retain fast reactions for which syntactic constructs imply scalar context.

So I leave in the crutch, and since everyone knows that I do this deliberately, I don't see any reason to comment it.

OTOH in code which will only be seen by people who are supposed to be competent Perl programmers, well I drop it...