@cupoftea: the pdf is not perfectly clear but the 10th light certainly does not draw 100 watt.
Ts states turning on an additional light with each step and notes in the table the theoretical, expected total power to the lights, ignoring cable losses.
I assumed that the noted smps voltage is measured at smps output, and that the 'power meter' is a mains device in between the smps and mains or maybe on the SMPS output, and it starts to make sense.
Any dmx bus is run at low voltage and means that each light contains some smart driver. Again I make an assumption, but it makes a lot of sense that each led driver contains a switching converter that feeds a constant current, thus a roughly constant power, to strings of leds.
@Eggdbdr1213: are these guesses right?
P.S.: ltspice can simulate this.. use a bv source and put p=10 in the value field. When the cable resistance is stepwise increased, a dramatic and sudden increase of load current shows above a certain cable resistance, maybe around 0.8 ohms (to be double checked).