Hi Ed,
Thanks for sharing your approach and Neal's response.  Sometimes we seem to forget that what we are trying to do is to convert the speed of the air to PRESSURE.  Your approach seems to be working well.  Have you ever measured the speed of the air moving though the radiator (where slower is better)?  I understand that this type of diffuser should reduce the speed of the air to somewhere between .1 to .4 of the freestream velocity, so I wonder how well your modified ducts work (I bet pretty well).  Also, do you know happen to know how well the wedge type duct (for radiators under the engine) recover pressure?  Should the wedge ducts also reduce the speed of the air to somewhere between .1 to .4 of the freestream velocity or they inherently less efficient?  If anyone else knows the answers to these questions, please chime in.

Thanks,
Ken Powell