...my weigh-in on the topic? ....
The higher the HP, the narrower the powerband, unless displacement is increased or carbs changed. You're right, Pin, that it gets harder to manage: for example, my 750 set up for 14,000 RPM yielded power from 13,500 to 14,200 and had to be geared to match the track where I was running: fast tracks got near-street gears and tiny tracks got 16T front and 64T rear.. If it's the street you're riding, you might want to get boots with thick soles in this configuration, because you'll be pushing off at those stopsigns..
As you widen the powerband, the HP will drop. But, if you put the power where you're riding, this often makes you ride faster. In the days of 1960s-70s Production Racing, this was the game: you changed the powerband to match the tracks by riding the track to figure out what RPMs you would need to handle the corners, then go tune for it. You were not allowed to change the tranny, but sprockets were OK. Cams could only be changed if the manufacturer sold the cam in their same-series street bikes (that's why I know a lot about K0-K4 cams
), likewise with pistons and carbs. So, installing a K4 cam (that had been "worn" to the production tolerance limits (-0.023") at the top and base circle of the lobes, using a grinder, but leave the rest alone) in a K1 engine would broaden the powerband from stock 6500-8500 to become about 5000-9000, with more bottom-end torque while preserving the top-end HP. This widened the cam lobe slightly, and the powerband goes with it. But, this gets into changing cams at the lobe. Wider lobes make more HP, reducing torque if the lobes overlap. If the overlap is kept down while the lobe widens, the torque stays there (this is why some "square" lobes appear in designs). If lift is increased (because the engine can benefit), the effective lobe width is usually wider by default. In the case of stock CB750K models, taller lobes have very little effect, but wider ones really help, because the throat is not a good breather in stock form.
The most successful street "drop-in" cams are those that start torque earlier with an earlier-opening, wider lobe, but keep the lift near stock for lower engine losses, with closed timing staying nearly the same. RC Engineering used to make these, but most people didn't understand their use. Every streeter I built with one of these became a neighborhood giant-killer. Chopper guys often went with the hottest, longest, highest cam they could, then found that they could hardly ride the bike in town. Chops, in particular, need low-end git-go.
The general rule-of-thumb for engine power here is the "area under the curve" concept, from calculus. Put simply, the graph you show above defines an area under the HP curve. If you image the curve being flexible, then "press down" the high RPM end, the power rises somewhere else to make the area stay the same, just like it was a balloon or something. (This presumes the carbs and cam and displacement stay the same.) Naturally, this is highly simplified, but here's an example: if you take the cam sprocket in a CB750 and slot the bolt holes where it bolts to the cam (about 1/2" both ways from center), you can "push" or "pull" the powerband around by advancing or retarding the cam by timing it in different places. If you advance the cam, the power comes on sooner, developing torque earlier, but giving up HP up high. If you go the other way, the torque is less down low, but they make more HP up high. Advancing 5 degrees on a CB750 would make the low-end torque much stronger at 4500 RPM, but top speed would drop about 5 MPH, gearing being the same. It also would feel "dead" above 7500 RPM. Setting it 8 degrees late would make the torque start about 7500 RPM and it would easily pass 10,500 RPM, pulling until the carbs got asthma. The "sweet spot", where the highest torque and highest HP both occur within the range of the oil supply and rod strength is generally where the manufacturer fixes a given cam design.
Although today, higher MPG is often a concern instead. That's a different trip.
