emanswfan said:

You use the 3D to create a depth map for one of the angles, and can then allow a lens blur tool (say in AE) to reference the depth map. You can then adjust focal point, and strength of the blur (how shallow you want it). you can even keyframe settings to simulate rack focus and stuff. It's extremely simple, but does require shot-by-shot conversion. It's like controling the camera settings after-the-fact.

The only thing to make it better would to use an algorithm specific to simulating depth of field, like field blur from photoshop cs6.

Quite interresting, thanks for the details :)