A friend recently bought a plasma table which was claimed to cut 30mm steel with no draft angle at all. Don't know how it's achieved there and haven't seen the machine yet.
You might be better using a lens with a longer focal length and further away from the workpiece, so that the beam travels more distance to the workpiece surface. May be that's what you meant...(?) But then you already have your design constraints. Also, can you play with the point where the beam is focused? Do you focus it on the surface of the material? Can you focus it in the middle of the material thickness?
The picture does have a nice effect indeed