The trouble with a lot of the pics people are posting is that the MAM manifolds are filled with unnecessary bends and a lot of extra length. The best manifold will direct air from the head to the turbo using the most direct path. There are several reasons to avoid bends and length:
1. Every bend disturbs the air flow - more bends, more disturbance. This causes flow reversions and eddy currents that will result in the engine having to work harder to push the air out and less energy getting to the turbo.
2. Bends result in hot spots in the manifold. Hot spots form due to the disturbances in air flow mentioned above. Every bit of heat lost is that much less energy being put into the turbo. Hot spots also result in cracking.
3. Unneeded length allows for more heat to be lost as the air has a longer path to take.
4. Extra length also results in higher EGTs which are hard on everything.
5. More manifold means more weight and weight is a bummer.
6. The more involved the manifold is the more difficult it will be to install.
The MAM parts are well made, look great, and have proven to flow well, but they have some short comings I'm not interested in copying. I've been rereading info from my turbo library (books like Maximum Boost, Turbo, and Supercharged!) for real info on the value of true equal length runners. The reality is that equal length runners aren't even mentioned. Conventional wisdom says that they are the way to go, and I would argue for them on an NA setup, but I'm beginning to wonder if they really matter too much with a turbo. The WRX is the only car I can think of where there is practical knowledge on the subject. They have very unequal length runners and this results in the Scooby rumble. After about an hour of reading I found this:
Speed Style and Sound Header Comparison There was surprisingly no discernable difference between equal length and unequal and this is on a car where the difference in runner length is measured in feet.
Final thoughts: After the last several hours of research and reading I believe that a difference of ~4" in runner length is completely negligible and if people like the look of the first iteration manifold thats what should be built.