Let’s start from the bottom:

(Assuming you mean a cannonball that is much denser, therefore heavier when the same size)

Initially, when they’re both moving slowly, so air resistance doesn’t matter, yes.

But once air resistance becomes important, the cannonball will go faster, so no.

There are three things going on: force of gravity (which is proportional to the mass), inertia (how much it resists acceleration, which is also proportional to mass) and the force of air resistance (which is more or less proportional to surface area, not mass). In a vacuum the cannonball has a larger gravity force, but also a larger inertia by exactly the same amount, so they fall at the same speed. But the air resistance force is the same for both balls; and because the glass sphere has less inertia it is slowed down more.

For rolling down a plank, it’s the same thing, except usually speeds are low enough that air resistance doesn’t matter much, so the cannonball will only be a tiny bit faster, probably undetectably so.

Now on to bicycles:

You’ve still got gravity, inertia and air resistance, plus some rolling resistance (friction on the wheels and bearings). Now rolling resistance is partly proportional to weight, and partly constant, so the lighter rider is affected a little more. And air resistance is important at hill-descending speeds, so the lighter rider is significantly more affected by air resistance.

Therefore the heavier rider has a speed advantage.

Plus, there’s the idea that a heavier rider might not only have a slight advantage descending but also a disadvantage climbing, so relatively speaking, descending is even better. Consider a rider that’s, well, let’s not mince mince words, he’s fat. He’s got the same amount of muscle for power as the skinny guy, but has to lift a lot more mass, so he’s much slower going uphill. Going downhill, though, gravity is more important that muscle power, so the fat guy has no disadvantage. And if fact a slight advantage, so he’s a lot better off descending. This works even if he’s not blubbery fat and out of shape, but just broad-shouldered with big arm muscles that don’t push the bike forward.

So putting these together, I think it’s not ridiculous to think that a smaller, lighter rider should plan to try and gain ground on uphill sections, while assuming he’ll lose some on downhill.