Of course it takes less energy to move a 1kg mass versus a 3kg mass. That's not what I was saying. I was saying the difference factor to roll a 3 kg mass versus a 1 kg is less than the difference factor of hand-carrying a 3kg mass versus a 1kg mass. "Simple physics" tells us that maintaining rolling only has to overcome the losses of drag and rolling resistance. If not for them, the conservation of momentum would continue to move either mass indefinitely. When walking, the only major conservation of momentum effect in play is the pendulum effect of your arms and legs swinging.

For every step, you have brief moments of acceleration, ergo why you move with a gait instead of a constant velocity, thus energy spent walking your body any distance is a linear function of distance and mass. When rolling, on a flat ground, if not for drag and rolling resistance, you only have to provide the initial acceleration to get up to speed and with nothing else, you would eventually get to your destination irrelevant of how far away it is, ergo the energy for frictionless rolling is not a linear function of distance and mass. Energy spent overcoming drag and rolling resistance is a linear function of distance. Rolling resistance is function of mass but it's only one component of it. You have suspension losses as well, but this is "simple physics", so we've approximated that out.

If not for frictional losses, energy spent to move mass is needed for acceleration not velocity because of the conservation of momentum. This is why it takes less total energy to move a man+bike a mile than a man walking a mile.