For humans, I think there's more to it than this article suggests. While it may very well be true that our most basic form of navigation is to remember exact visual snapshots, it seems clear that we are also able to augment this ability with other mental tools (e.g., Songlines).
I read the beginning of this thread before going on a 32-mile overnighter to a completely new area this weekend, so I consciously examined how I was making route finding decisions and generally finding my way around. I found that reading and constructing my own "cognitive map" of the area was a process of (1) translation of the printed map into actual views and (2) translating and relating my actual views back into three-dimensional space.
(1) I see an area on the map where the trail starts on the left of a small southern hump on the ridgeline, passes between it and a slightly more northern small hump, and proceeds along the right side of the northern hump. The generalized landcover shown on the map (a USGS quad) indicates that there aren't any trees in this area. Looking to the east, I should be able to see the drainage of a large, prominent creek. There are large waterfalls along it, so, if it's quiet, I should be able to hear the sound of falling water. Thinking about the area in context and relating it to my previous experience, I can guess it will be exposed and rocky and feature a certain collection of sub-alpine meadow wildflowers. Taking it further, I might even expect to find and hear certain kinds of birds and insects there. So, I put this all together in my mind and translate it into the sensory experience of standing in and traveling through that place: what I will see, hear, feel, touch (and in some cases taste if I'm specifically looking for food or water).
(2) Standing at a viewpoint, I look to my left and right and see two creek drainages. I look at the overall changes in forest canopy relative to elevation, using what I know about undergrowth and wildlife in respect to elevation to fill in the gaps of what I can't see underneath. I estimate the slope. I look for outcroppings along the creeks, listen for falling water, determine which direction the water will be flowing etc. I take all these things in, consciously orient everything in three-dimensional space, and add it to an overall mental map of the area that I can then later use the techniques of (1) to translate back into located experiences.
I would consider these two inter-related processes to be learned skills that I can consciously apply "on top of" my underlying, built-in visual snapshot navigation method and can further improve with practice.
For an example of a different kind of learned mental skill that can be used to augment our built-in abilities, check out this article: "Secrets of a Mind-Gamer" by Joshua Foer.