1. Yes, you still have to move the rest of the rotating parts in the car... that's why we try to buy things like lighter pistons, lighter rods, lower loss transmission components, and lighter wheels/tires/brakes. You try to gain every little bit where you can. I really don't know how much more power a lighter flywheel would free up, but you could do the math figuring in a reduction in rotational intertia from the loss of a few pounds on a rotating disc. I don't really want to sit down and do the calculations right now, not for this thread. The main advantage I get isn't in faster acceleration of the car- it's drivability on the track. My engine revs up and down faster now, so I can rev it up faster when I'm at the line waiting for an instant green at the track, or I can shift faster on the road course without worrying about the heavy flywheel jerking me around when I engage the clutch. I personally couldn't tell much when I put my new flywheel in, but I get used to changes in driving dynamics so quickly that it doesn't matter. I also changed a lot more than just the flywheel when it went in, so it was hard for me to judge what stemmed from the flywheel change and what didn't. By the way- lighter components don't really add power- they just
reduce the amount of losses through your drivetrain. Your engine isn't making any more power than before... more power is just making it to the wheels than before.
You're not making more power- you're just losing less of it. Just a technicality...
2. The drums on a standard Dynojet dynamometer are set to simulate a 3000 pound car at default, I think... so when your Honda is tied down onto one, it's actually trying to move more mass than it does on the street. I could be wrong on the preset mass simulation with the rollers, but I think that's the right number. You're not moving 2300 pounds of car anymore- when you're on the dyno, the engine is trying to move 3000 pounds of
simulated car. The power increase at the wheels that you measure are pretty close to what you would feel in the car.
As a side note, it's hard to correctly tune a car to 100% on a chassis dyno with fixed rotational inertia on the rollers. If you're not simulating the proper mass with the rollers, you end up tuning for the wrong air/fuel ratio. Rollers set to a simulate a weight/mass heavier than the car strapped onto them will make the engine pull a higher load than it would on the street, and lighter rollers will do the same thing- just in the opposite direction. The best way to dyno tune a car is to use a dyno with adjustable roller mass or adjustable weight placement in those rollers to change the rotational inertia and simulate the proper weight of the car for the engine to pull against. If this isn't available, you can tune to the best power on a fixed weight roller dyno, find out what air/fuel ratios your engine makes the best power at for each given rev band, then take it out on the street with a wideband sensor and tune it again to the max power a/f ratios. This way you're tuned for max power pulling against the exact right load, and you'll get the best performance. On the other hand, tuning against 3000 pounds of simulated resistance really wouldn't be that bad in our Hondas anyway, since you probably weigh about 2700 pounds total on the street, and then you have to deal with wind resistance too, and maybe a few heavy things in your trunk. 3000 pounds is close enough for me.