Perhaps it’s not always a new heuristic that is needed, but a better way to manage the externalities around current/preceding heuristics.
From a “knowledge-searching” perspective, at a very rudimentary level, it makes sense to look to sites/pages that are often cited (linked to) by others as better sources to rank higher up in the search results. It’s a similar concept to looking at how often academic papers are cited to judge how “prominent” of a resource they are on a particular topic.
However, as with academia, even though this system could work pretty well for a long time at its best (science has come a long way over hundreds of years of publications), that doesn’t mean it’s perfect. There’s interference that could be done to skew results in one’s favor, there’s funneling money into pseudoscience to turn into citable sources, there’s leveraging connections and credibility for individual gain, - the list goes on.
The heuristic itself in not innately the problem. The incentive system that exists for people to use the heuristic in their favor creates the issue. Because even if a new heuristic emerges, as long as the incentive system exists, people will just alter course to try to be a forerunner in the “new” system to grab as big a slice of the pie while they can.
That’s a tough nut for google (or anyone) to crack. As a company, how could they actually curate, maintain, and evaluate the entire internet on a personal level while pursuing profitability? That seems near impossible. Wikipedia does a pretty damn good job at managing their knowledge base as a nonprofit, but even then they are always capped by amount of donations.
It’s hard to keep the “shit-level” low on search results when pretty much anyone, anywhere, anytime could be adding in more information to the corpus and influencing the algorithms to alter the outcomes. It gets to a point where getting what you need is like finding a needle in a haystack that the farmer got paid for putting there.