I would prefer to wire in parallel and use berg or molex connectors to hook up to the lights if this is going to be a yearly setup. If the light bulbs are incandescent then polarity of the wires won't matter. If they are led lights then use a red sharpie or paint pen and color one wire to indicate the polarity.
With series wiring each bulb will drop (or consume) a certain voltage. The amount of voltage consumed by each bulb will determine their individual brightness. The three lights will divide the total voltage according to the drain/load of each light. Given the quality of today's toy products, the variance in the internal wire, the bulb, or internal series resistors would all have an influence from light to light. The size of wire that you use to connect them to your power source could have an influence also.
With parallel wiring, each light will get roughly the same voltage. The voltage would decrease down the wiring according to the size of your parallel wires used. 26 or 28 gauge would possibly be noticeable while 12 gauge wire would not be noticeable.
You could lay three passenger cars on their side and connect them with clip leads in series going from pickup rollers of one car to the axle of the next car to see if you have any variation in brightness. If you put them on the track and connect your power source to the rails then they would be in parallel. It should take more voltage to light the cars in series than in parallel.
