The length of wheel spin has little to do with the bearings, which in fairness, may be the least contributing factor.

First of all new bearings (new hubs, but keeping it general) are more fully packed with grease than they'll retain in service. As you ride the balls push the grease into berms on either side of the track whereby they pick up only a film as they pass by. Grease is highly viscous, so pushing a pack of grease around will make any bearing feel sluggish.

There's also seal drag which hubs without seals don't have, and in those with seals the drag will be reduced with wear.

So you have two good reasons why a new bearing will have more drag than the same one later on.

Now consider the wheel, not only is there hub drag, there's also the inertia based on the weight of the rim and tire, and the aerodynamic drag of the spokes. These are important, because hub drag is a fixed constant applying a given braking force, so the wheel with higher inertia will spin longer (all other things being equal). Aerodynamic drag will drop as the wheel slows, but never goes away completely and can't be discounted.

ow looking at the big picture, the forces involved in how long a wheel spins in your test are miniscule in the scheme of things, (unless something is Very wrong), so even a seemingly large difference is meaningless when considering a bike and rider weighing shy of 200#s. The two factors that most determine drag on a rolling bike are aerodynamics, and tire drag. Small changes in body position or tire pressure can dwarf the seemingly large disparity in the hubs.