A mixture really Al, there are some exacting sciences behind it being succesful, you're quite right, but its more the application of what parts of the science are needed for the job at hand.
For example, you mention Valve Size/Area, it could be huge, but then if in relative terms the valve only opens every other lampost, it becomes difficult to use the answer within the design because nothing else would fit in.
the biggest difference that more valves/area, i.e the 16 over 8 would make to the overall exhaust design, would be that the 16v is designed to rev reliably higher, and work generally more efficiently at higher rpm, but the truth of it that in our boxers, the difference in the way that the engines behave and ultimately perform isn't that big in the end result. Yes the 16v has more valve area, but also, the 16v doesn't now rev 3k more than the 8v, and the 8v's combustable route, swirl, flame pattern etc works well across the rev range anyway, so in terms of changing the exhaust design required, i would say no, their simply isn't that much difference, and an 8v should respond to this exhaust pretty much just as well.
out and out engine size plays a large part, and instantly points toward certain size parameters, but then so does cam profile, compression, and the overall efficiency of the combustion swirl!
not all of these are taken into exacting consideration on a daily basis i'll admit, like many trades, experience cuts out a lot of math, however, the beauty of this design meant that we had some very useful figures in the form of proven sizes from Dave Ashford and his race engines, and Rodgers ANSA crossover, which although not perfect, ANSA were no mugs, back in the day they would have put a lot of R & D into that, so it was an excellent starting point.
Daves a hero though, trust me, its no mistake the Rodger is having his race car fully prepped by him.
believe it or not, boxer layout it may be, but it shares a lot of mathematical and behavioural characteristics with the twin cams, which means we can plan the sizing as if it were a normal in line four cylinder, almost the same as we would an early 2.0 twin cam or as late as a 2.0 TS, and then transpose that into the boxer layout as best we can.
i say as best we can, because ANY exhaust is a compromise, no one exhaust is efficient throughout a large rev range, the fewer the revs, the easier it would become, so something as freely revvable as our boxers, does present something of a difficulty.
shorter headers/primaries are great for top end power, but the shorter they are, the more you're going to lose at bottom and middle. having a 4-2-1 design, instead of a 4-1, means that the secondaries can fool the engine a little, so you can run shorter headers for the top end, but the secondaries still provide some additional length for mid range, get these both right, and you're on your way to a winner.
in the ideal world, you would have a varying length exhaust, so that as the revs increase, the primaries shorten, and this has been experimented with many times. The bike game is probably the closest with a reasonable affordable solution. Yamaha started it with the EXUP Valve, which is nothing more than a butterly valve that sits in the single collector pipe. At low revs, its is partially closed, and the more revs the more open, until eventually by about 6k they are fully open to let the short headers do their top end work alone. It simply creates more back pressure, but its one way of fooling the engine into behaving as if it has long headers at the bottom, medium headers in the middle, and it actually has short headers for the top. And it works, take this out and you end up with a big hole in the midrange, but the bike feels quick because the blob of power came in with a surge as the short headers found their sweet spot! Funny how many people feel for that one!
as for sizes, the stock side by side layout is much bigger Internal Diameter, but look at the length of them too, its remarkable that AR fitted them at all, and the biggest reasons for them is the compromised space allowed, price, and ease.
our sizes, whilst closely mirroring the ANSA R & D findings, and Daves wishlist for a Dream Pipe, were adapted into our own crossover, but the lengths had to be compromised from the paper dream, to the space provided, and the route to be taken therein. Thats life! but we improved substantially on the ANSA attempt at equal length, so our scavenge should be much better, and negative resonances are reduced, i.e. it is much better balanced.
that leads us into gas harmonics, but on a day to day basis, with road or road based engines, gas speed, flow, and extraction are the main concerns, so the right sizes, in a well formed shape as possible within the design allowances, will flow well, so if its well balanced too, then the rest of the system demands, such as back pressure, scavenge, extraction etc will all also come along for the ride!
is that what you were after Al?
i can measure the original layout when i'm back over at the workshop if you like, i don't have it in front of me.
phew, tired out after that