The Honda engine started life with the NR500, as you say, to circumvent the rules limiting number of cylinders. Apparently this thing revved to an unheard of 22000rpm in 1979! As a concept, brilliant - but a complete disaster as a race bike. This thing even had a monocoque chassis and I seem to remember the fuel tank below the engine. Perhaps a classic case of losing sight of the bigger picture and just seeing the engine in isolation?
I wonder why desmo valve gear never really caught on with racecar engines? I can see why not these days, but in earlier times
"5.1.4 All engines must have 10 cylinders and the normal section
of each cylinder must be circular.
5.1.5 Engines may have no more than 5 valves per cylinder."
As for the Honda NR750, the oval piston technology was developed to skirt the GP bike rules limiting 4-stroke engines to 4 cylinders. The "4 cylinder, 32 valve" oval piston engine, was in effect an 8 cylinder, 4 stroke GP engine. It may have had only 4 pistons, but it also had 8 conrods, 32 valves and 8 spark plugs.
The issue of the optimum number of valves per cylinder, for a N/A F1 engine has been thoroughly vetted. Obviously, a 4-valve, narrow angle, pent roof arrangement is best, because that is what everyone uses. F1 engine designers aren't stupid. If a 5-valve configuration was better, that is what they would use.
In F1, it's a case of "monkey-see-monkey-do" and Darwinian "survival of the fittest". If a particular design feature (engine, suspension, aerodynamic, etc.) demonstrates an advantage, everybody else soon copies it.
On paper and in theoretical parametric studies, the 5-valve configuration always outperforms the 4-valve head for high-speed racing engines. In practice, however, the 4-valve configuration performs just as well, using 25% fewer parts.
Think about this: Cosworth currently has a 32-valve, 2.4L V8 F1 motor (4-valve/cyl) running at over 20,000 rpm. End of argument.......
