Coming from someone who builds exhaust systems for a living that is a real compliment, thank you!
you haven't eliminated the bottom end torque peak, just moved it around a bit, and maybe softened its arrival
Is that the case here? I wasn't sure, because I had worked out how the system was working but sort of empirically. My logic went along the lines of, if the primary lengths are optimised to provide peak torque at a certain rpm, dictated by their length and diameter, then the proportionately longer path connecting opposing cylinder pairs (4 and 1, and 3 and 2) would result in a secondary torque peak, but lower down the rpm range because of the longer effective pipe length. It followed that the shorter the secondary pipes were, the closer the two torque peaks would be in rpm, until you get to the 4-1 system favoured by the racing guys where all the torque peaks at the same rpm.
That led me to conclude that, if based purely on effective scavanging of the cylinders, only cylinder pairs 1 & 3, and 2 & 4 are tuned for peak effectiveness, and in order to correctly reinforce the secondary torque peak, both secondary pipe lengths would need to be the same, otherwise you would lose the benefit of the scavenging pulse on cylinders 3 & 2, and 4 & 1 (it would arrive too early to be effective for one cylinder and/or too late for the other).
So my assumption was that any benefit from the secondary torque peak would be lost altogether if the secondaries were unequal length, but as you say, there is probably a broad operating range where some benefit is being felt, so maybe its not as bad as I thought.
My old exhaust was based on the original 1.5 single carb downpipes (1" bore), extended to more than twice their length(!!!) before going into an EMPI VW silencer (!!!???!!!), which looked like it had a crossover between the two silencers, but in fact didn't!
Anything was going to be an improvement on that!