You need to read up on how SEMs do their imaging... CB is not clear on it but probably easy to find on Wikipedia say. As CB recalls, the machine takes electrons emitted from the surface and focusses them back to form an image. So really this is an "emission and focussing" function.
The image thus depends on how the electrons are emitted as a function of direction. What we see in SEM images suggests something like the "rough surface" reflection function, in which a rough surface like the moon's bounces rays around and ultimately re-emits them uniformly in all directions.
Remember, Lambertian surfaces look the same from all directions just because they do NOT emit uniformly in all directions! Foreshortened surface areas cancel out the non-uniform emission.
It's a 5 minute exam question to say why, therefore, the full moon looks "flat" instead of rounded like a ping-pong ball. Also one can ask the 5-minute question: "what does a rough sphere look like in ambient light?", and the maybe surprising answer is that it's brighter around the edges and dimmest in the center. Convince yourself.
Not unlike this example SEM image, for instance: