[Fred68]

Manifold and exhaust design is extremely complex. In reality, to get optimum torque from any engine requires a custom designed system. A optimised manifold can help the cylinders fill with the next air charge. Now obviously anyone building systems for open market would need to compromise the system. Manifold primary and secondary length is dictated by many factors. The size of the pipes is very important. Too larger and gas speed will be slow. Too small and although gas speed will be high, back pressure will increase.
Primary length is dictated by many factors. Things like bore diameter and stroke, rod length and valve open/ overlap period are all important. To make things more complex. The optimum primary length is only beneficial at certain points in the rev range. The trick is to get benefits in key points in the rev range at the same time, not causing massive losses elsewhere in the rev range.
This is where secondary length helps. The secondaries can be optimised to lift low points in the rev range.
If all the science is removed from the design. The very best option is remove anything in the manifold and system that causes a restriction to gas flow.
This is why manifolds like the JS manifold work. It's not adding power to the engine, but it doesn't rob to much ether.

Years ago, I made several manifolds with slightly different primary lengths. The theory was to spread a smaller power increase over a wider rev range. It worked too but I found with this method, carburettor calibration almost impossible. Mostly because anything done to the exhaust is reflected in the intake.