Indeed, this "theory" seems to involve talking some particular "thing" the brain "does" very prematurely, at the level where the speaker just knows too little about the brain's internal model to make their concept even meaningful.

Article: "At the time, the dominant theory in psychology for how people find their way was the cognitive map, which posits that humans and many animals create representations of the environment in the brain that they use to navigate the world."

As you say, pretty much tautologically, an organism that navigates its environment has a model of its environment. But to make the above statement falsifiable, the theorists apparently put the theory as "a uniformly coded model existing in a particular spot in the brain", which is a big, big step.

The thing is, if one has ever worked on a large computer system, one knows that just because a program shows some functionality, that doesn't mean that there will be discreet module separable from other other modules that implement this functionality (there might be such a module but it's functions are likely wrapped with many other modules). Yet it seems like cognitive scientists reason in this constantly.

No but there's probably a specific part of the brain that does deal with visual mapping.

I read this excellent article some years back about the test that aspiring London cabbies, the ones that drive the black cars, take and the preparation required "The Knowledge" to pass this difficult test. Since London's road system is so byzantine and requires lots of insider knowledge, it is unlikely that the cabbies will in the near future ever be supplanted by companies like Uber or Lyft.

Anyway turns out the gray matter in the posterior and anterior hippocampus of some of these guys were tested before and after studying for the test and there was significant growth. Perhaps this would be the place to start to pin down a neural equivalent to a real-world map.

https://www.wired.com/2011/12/london-taxi-driver-memory/

Sure! In fact, you could say that the eyes are the part of the brain that has a close isomorphism to (a tiny subset of the EM manifestation of) the real world. Once you get about 2-3 cm from the retinas, though, I think that information has been encoded in ways that make it unrecognizable from the outside. I'm saying I doubt we'll ever be able to decode a model (let alone a map) of London from studying a cab driver's hippocampus.

I have to adjust me "priors", as a Bayesian might say. Still, monkeys aren't humans, and faces aren't London. Recognizing faces and their expressions has been a thing since faces first evolved, eh?