Generic stream of babble not meant for anyone in particular, but feel free to be offended:
I have to wonder if the old cam profile the JS04 is based on was ever used with a Norton head that had the reduced exit D exhaust ports. If the D port concept on an old twin head is as good as people selling it claim, why would a 1 1/2" exhaust of the correct length make any significant difference? I seriously doubt anyone making significant HP with a Norton engine aimed at racing on the salt flats is running 1 3/8" diameter exhaust pipes. Then again are people talking I.D. or O.D.?
I too will add to the stream of babble, and lend my amatuer thoughts.
I have talked to Jim Dour at Megacycle about exactly this. The Original Norris/Sifton 480 was created for a Triumph, and D ports were not common on Triumphs in that era. In those days both Norris and Sifton made a profile and adapted it to everything they could. The cams existence was to push the limits of mostly lift and tappet strain on a Triumph. This was all done long before computer cam design was available to independent cam grinders.The closest thing to computer design in those days was Jim Dour would go to the Boeing offices in San Francisco after hours, programming punch cards and then take the feed sheet print outs and use a dividing head on a mill with a dial indicator that read in ten thousands and subdivide those lines by eye in hundred thousandths to make cams for Dick Mann's A65 and A70 race bikes.
Needless to say the JS4 cam has been significantly been reworked by Jim Schmidt since the Norris days and it is a very different cam, the lift is probably the only thing that is similar.
I am probably over simplifying this, but with the pipes the larger OD of the pipe will effectively create more duration. The pipe will change how long the cylinder keeps moving exhaust after the valve opens, and how much it pulls in during overlap. The pipe helps to create a time area for the escaping gases and the time area is slower with a larger pipe and faster with a smaller pipe. It does not change pressure wave as much, but it does change velocity of the gas. At lower RPM, the pulse can be slower. But at higher RPM, when the engine is moving a lot more gas, the larger pipe can keep flowing after the smaller pipe would start becoming a restriction. A smaller pipe is usually easier to manage as the exhaust gas velocity helps minimize the effects of reversion. However, it is limiting as well in other ways, which gets us into the nuance of cross over pipes, x pipes, merge collectors, etc.
If the excessive crankcase pressure issue blowing out the main seal is not resolved the engine will be up against a compressed air wall on piston downstroke and never produce what it is capable of.
The question(s) I have is, did the cam/valve size/port size create an airflow mismatch that made more cylinder pressure leak past the rings? Are the rings are fluttering or is this detonation? Is the lack of air flow or port velocity creating reversion with at high lifts and working past the rings? If so is this intake or exhaust reversion or both? My personal suspicion is that a lot of exhaust gas is pushing back into the chamber during overlap as velocity is lost due to inadequate flow at high valve lifts. But this still may only be a symptom and not be a root cause, there could be too much windage for this much pressure, or the reed valve is not large enough for the needs of this engine, a bad seal, or a million other things. But this is the optimize the system part of building a race bike.
The JS02 I had in my engine for a while was strong and felt like it did not run out of legs, but I'll never know how much power it made because my bike has never been on a dyno. The JS02 didn't feel any stronger than my old SS cam though. I actually prefer the SS because it is a little more violent when it comes on. It also works a little better on the street with a lighter crank. but i digress
Also advance your JS2 a degree or two and compare it the SS cam, I would be curious to hear your thoughts on the comparison.