I interpreted Graham’s post to mean he would be willing to take a shot at producing an example, using the OP’s own source material as a starting point. Maybe someone else might not understand his post the same way.
Given the difficulty sometimes of interpreting what people really mean in forum questions and answers, it seems the “How could anyone …” response is inappropriate. Imagine if someone were to answer “what’s a still” by saying “How could anyone on this forum not know what a still is?”
As far as a moving camera point of view is concerned, that requirement has not been spelled out at all. With a background still and a green screen foreground, your tricks come from panning, zooming, tilting (distorting a still to simulate a change of perspective) and relative motion of fore and back. Fancier tricks might come from lighting or focus variation. Even fancier tricks come from carving up an image so that different parts are panned-zoomed-tilted individually.
How far is someone willing to go to simulate camera movement? If the original still is not wide enough, some people might go to some photoshop effort to build out the image to one or both sides using elements already in the scene or elements taken from other photos shot near the same location at the same time. Sometimes the best photo compositions are not the best source material for video just because they leave no room for panning or zooming.