Seems obvious now, but make sure the camera is completely within the bounding area of your confiner polygon. That is, the camera should be smaller than the confiner. It wasn’t obvious in the video, but the shape of the virtual camera will “hit” against the sides of the confiner area to block it.
When I set up my confiner polygon originally, I wasn’t watching the shape of the virtual camera’s view in the Scene View so my confiner shape was taller than camera but the camera was wider. The results was that the camera never had any bounds to hit (since the bounds were already within it).
I hope that makes sense and helps someone else that made my mistake - I couldn’t find this solution online and finally realized what was going on by just playing with the shapes.