In this picture it looks like the opposite of anti reversion in the header ports. Looks like the exhaust pulse hits the sharp edge of the primary pipe. Almost like they wanted to create non laminar flow.

 

This is a picture of a NON OEM header, this design seems to make more sense then the oem recessed primary pipe. It's hard to tell from the picture of the OEM header how far set back the primary tube is though.

 

None of the exhaust makes sense, they fundamentally all contribute to the loss of exhaust velocity to soon in the exhaust system which may increase the MPG and possibly lower emission but also rob the engine of making and sustaining torque a little longer. You can see this in many platforms not only Mazda by data logging the CAT temperatures. In the few platform I was able to use EGT at 3 places in the exhaust, primary header, mid pipe or after all CATS and close to the tail pipe. It is easily 35-45%+ drop in exhaust temperature by mid pipe.

 

So are you going to get your die grinder out and take your header off? Do some port matching?

Have you ever taken your header off? Is it difficult to remove?

Some people were talking about having to drop the sub frame to get it out.

 

doing some more fishing...... LOL
Most intake or head porting on anything these days I leave that task to my friend and business pardoner that has a Performance Head Porting shop/service in FLA. I will however do all the final polishing if needed. But I do the testing on an air flow program and give him the spec's that we need to get to . When it goes that far not using stock valve train parts either.