If the situational interpretation applies, then reducing the use of semantic stopsigns would mean less available cognitive power to respond to meaningful counter-evidence, not more.

I think situation plays a role here as well though.

If I'm reading the comments section on Shakesville and see some rando come in with a basic question and get hit with the "I'm not your sherpa" card and a link to 101 materials, that's fine. You can't drop everything to debate every random dude who expresses a disagreement; I certainly don't appreciate it when people wander into the bio department and start up debates about irreducible complexity (yup, true story).

On the other hand, if I'm on GiantITP having a fun conversation about the best way to generate ability scores in Dungeons and Dragons (3d6 down the line, BTW) and someone goes full RadFem and derails the thread into talking about "biotruth" and privilege until it has to be locked, my jimmies get considerably rustled. Especially when I recognize a lot of the same rhetorical techniques I saw up in the first example.

That's the general point I was making; these tools are useful for defense, but unfortunately just as useful for offense.