Originally posted by Bosswrench: FWIW, engine builder & author David Vizard has written about headers several times in his various books, and according to his dyno tests, if the individual tube lengths are within about 6" or so, no power gains/losses will be seen.
Actually, if you are referring to his writings in his “Building Horsepower” series, when Vizard made this statement it was specifically in reference to dual plane crank V8s and conventional 4-1 headers. This is due to the asymmetry in how the exhaust pulse phasing arrives at the collector in a dual plane V-8. Though dual plane V8s with 4-1s are relatively insensitive to primary length, they can be very sensitive to collector diameter and collector length. In fact, collector length can be a very (if not the most) effective parameter in tuning such a system once the in car constraints and layout is fixed. When you go to 180 degree exhaust in the same V8, equal length primaries certainly does matter. However, this whole discussion must be footnoted with the fact that it is most often the case that it is impossible to achieve the theoretically optimal primary length because you physically cannot make equal length primaries reach the collectors without adding additional length. This is certainly true in Pantera’s and GT40s. The most popular embodiment for dual plane V8s remains the 4-2-1 or the Tri-Y, as most commonly seen in NASCAR. When done properly, they are very long for Pantera fitment though possible. The good news is you can still build a high performing zero-loss exhaust system within the constraints of a Pantera.
What's more important is the fit of the individual tubes into the weldment. Sloppy tube fits buttered up with weld do not work as well as neat fitting 'fishmouth' tubes, and often crack. Further, mild steel tubes will expand and contract by as much as 0.060", and stainless even more, which is really what cracks poorly made headers.
In the same articles, Vizard also comments that 4-1 Dual Plane V8s are remarkable insensitive to dents in primary tubes, with sometimes no noticeable performance effect with even 60% of the diameter compromised. So all you sledge hammer mechanics rejoice.