You must first devise a simplified model for "walking". One example, that might seem ridiculous at first, could be a rolling polygon, like this:
View attachment 266494
Take an ##n## sided polygon. How much energy is required to roll it around to the highest point (standing vertically on its edge), assuming it will have some more GPE (how much more?) and some amount of kinetic energy at that top position?
Do we assume that all of that energy is dissipated away as heat when it lands on the next flat surface, and this cycle continues as it rolls? How might you adapt the model for different types of people?
This is just wrong. The first ball has kinetic energy ##\frac{1}{2}mU^2##, the second ball has kinetic energy ##\frac{1}{2}m(2U)^2##. I have no idea what you mean by energy "spent". If the surface is smooth then neither undergoes a change in kinetic energy.