Recognizing that any discussion of changes to board behavior is moot and we’re just wasting electrons. IMO …
Admittedly GQ has a very different “mission” than the other forums. Though the true mission, even in GQ, is for folks to have fun reading and posting and for TPTB to make money off showing ads to those folks.
Solely within GQ we could (in theory) have something very different from a like button. We could have expertise badges for various specialties. Or we could use the “report post” function for flat erroneous posts and a mod could add a note: “This post disputed as to factual accuracy”. All of which drives in the direction of wikipiedia, rather than a general interest MB. Notice that wiki works because the amount of curation is significant compared to the amount of authorship. IOW, that approach won’t work here.
Another approach, the one currently in use, is that when a question comes up about e.g. physics I look for Chronos, septimus, Pasta, and a couple of others to weigh in. Knowing that those folks know the real deal and also have the expertise to introduce the right amount of simplification and footnotes. Everything else by everyone else (including by me) I treat as suspect.
Which amounts to me creating my own set of “expertise badges” that I award as I will. Which is of course only as good as my own ability & inclination to identify good vs. bad answers.
Ultimately any system based on the audience up/down voting any aspect of a post is only as good as the audience. Every attempt at self-curating systems descends to the level the audience is willing/able to maintain it to. For stuff like advice on cooking, most folks are knowledgable enough that their collective assessments wouldn’t be too far off. For something like particle physics, not so much.
My bottom line: You’ve identified an actual issue that in an ideal world could & would be corrected. IMO that’s admirable, but out of the board’s scope, and certainly beyond its practical resolution. A system which leaned in that direction but appeared to promise far more than it actually delivers is worse than useless. e.g. xkcd: TornadoGuard & xkcd: Star Ratings