Render target displayed on textures would probably be the easiest way.
So you have your mesh ( or whatever ) in one location, with a camera trained on that. The camera sends it’s output to the render target, and the you have 3 objects in the other 3 locations which have the render target as a texture.