This can result for some relatively detailed results without having to spend huge amounts of time modelling complex geometry. Of course, the method comes with its own challenges mainly having to create and project several variations of the image to cover any loose spots depending on the movement of the camera.
Some more could be done to introduce more depth into the scene with appropiate scene dressing and camera work. But as a quick experiment, I'm relatively pleased with the results having carried out the process myself, and understanding how to better approach this method in future.