Here's some more exhaust talk.
I just reviewed some of my books on exhaust system design. Although there are formulas for just about every aspect of exhaust system design...that tell you how gas behaves in the system...you would be amazed how often these books, with all their science tell you that a truely effective design is often a matter of trial and error. There are a few hard and fast rules concerning how primary pipe ID and length effect power (torque) production at a given rpm range. But when it comes to hard and fast rules beyond the primary ( or straight diameter pipe) they just don't seem to exist. The theory behind megaphones is they basically slow the pressure wave, which prolongs the scavenging effect...picture it this way...if you can slow the wave and fit more waves heading out the pipe into the meg then their tendency to pull more gas out from the primary pipe is increased. However the down side is more surface area of gas to pipe...which produces friction and causes the waves to lose energy. All testing has shown that a properly designed megaphone will increase engine efficiency over a straight pipe...but a poorly designed meg will hurt power and produce less than a straight pipe. Ideally a meg causes the exhaust to give up most of it's energy trying to pull gasses out behind it...meaning a perfectly designed meg would probably have a fairly soft even flow of gas out the back rather than strong, pulses that knock your hat off ten feet away. Remember...the energy to produce those hair fluffing pulses ten feet from the meg has to come from somewhere...it comes from the engine...power that you'd rather put to use scavenging exhaust rather than blowing peoples hair around.
The final opinion of a book full of formulas is that megaphones are for the most part best designed by trail and error (with reasonable starting points of course). And when it comes to reverse cones...it says the resulting gas flow becomes so comples that there is no formula or theory that seems to encompass the wave dynamics produced and that there is no evidence that it produces more power than a properly proportioned megaphone. It may however correct and issue with a improporly proportioned meg.
In general the same old guidelines apply...short primary pipes and short megs enhance high rpm power ( because the higher frequency pulses need less length of pipe to fit in) and longer primary pipes and megs enhance lower rpm power (less frequent pulses need a longer pipe to fit the same number in).
But then...what are you designing for? A dyno test or a race? A short system may provide higher peak hp numbers at the upper range of rpm...but produce poor power at low rpm. So for instance...take a high reving 200cc engine....it has very little power at lower rpm ranges....so, dispite the fact that it's a high revving engine...you put a long system on it to help it where it needs the most help...in the lower rpm ranges. The longer system hurts high rpm power less than it helps low end power....for a net gain of area under the hp/rpm curve. Which should always be your goal...not the hp peak, but the area under the curve...the total area under the curve represents your usuable power during a race.
Strangely enough in the old days it was popular to run straight pipes about 8" long ( look at some old photos of bonniville machines). Why? because very short primary pipes don't help make power...but...they also do nothing to hurt power...so at the least the tuners knew where they stood as far as the exhaust system went, and they could concentrate on other aspects of engine design.