Quite right. Good catch.
I was mistakenly thinking all the energy of, say, a show absorbing footfall went to heat. That’s wrong.
Well, a lot does go to heat. Consider a person running a few miles on a straight level course, for instance: They obviously have to spend some energy to get up to speed, but then once they’re at speed they need to continue spending energy to keep moving. All of that energy is getting converted to heat somewhere, because neither the kinetic nor potential energy is increasing.
In practice, though, I doubt that much of that energy that goes to heat can actually be reclaimed.
I wonder: if our jogger is running through room-temperature, say 20C air, and is wearing a special headband to convert temperature differentials to electrical power, how many watts can be extracted?
There are some companies active in power generation systems that harness footfall. Here is one company that seems to be making a noise about it.
However, whether the numbers add up, that is another matter. A lot of the projects seem to be demonstration projects with an educational intent or for promotion or public relations.
If you have access to a reliable electricity grid these systems have few advantages unless they were built to scale in areas of high footfall. Digging up the ground is never cheap in urban areas. But if you were in place that has no reliable electricity grid, such micro generations makes more sense. But then those parts of the world would be hard pressed to afford the capital cost of the technology.
Nonetheless, with cheap LED lighting the improvements in battery technology, this may change. There is also growing demand for powering sensors - the Internet of Things - collecting data about the environment and collecting it using some kind of wifi. These little things need power and it is not feasible to send someone to change the batteries, so energy harvesting systems are important area of research. For powering floor based sensors, it makes sense. They would probably make more money out of the data than the electricity.
I would be interested to know the numbers add up. This gives an indication how little energy is generated by a footstep. Not enough power to charge a smartphone, but certainly enough to a sensor.
Bulk generation of electricity for power hungry appliances is a very different problem from microgeneration and powering lots small sensors that may provide a flood of useful data.On a more domestic and practical level, google ‘kinetic switch’, these gadgets make a lot of sense in the home. Powered by the energy of you switching it on and off, it is connected by radio to receiver that can switch a light or any other appliance. No batteries, a switch with no wires.
Generating tiny amounts of power for single functions like sending a switching signal, can be very useful.
Usually forgotten in this is the energy available from the hydraulic pressure of expelling urine. Research was done on this by Richard Feynmam in the 1930, when still a college student. And this energy would be fairly easy to harvest, by connecting a catheter to a miniature hydroelectric generator.
I’m gonna bet this research was done while Feynman was age 20, not 70. A lot less hydraulic pressure is available later. 
Feynman didn’t live to be 70…
That puts our hot reservoir at 310 K and our cold reservoir at 293K. So the maximum thermodynamic efficiency is (310-293)/310 = 5.5%. Typical power expenditure of a human is about 100-200 W, so we’d be looking at a maximum of 5-10 W harvestable. And any way of doing that would probably interfere with human temperature regulation, and thus be dangerous.
Neither have I … yet. But I hear tell about that pesky hydraulic failure.
The elastic recoil bit however does not apply to clothing in which the electricity generating threads are not in addition to threads (and the deformation of them) already present but in replacement of them. Whatever stiffness a bit of clothing already has the hypothetical bit of clothing could have, but with the deformation of the material generation small amounts of power instead of heat.
And I get the point about elastic recoil in shoes but much of the energy of each footfall is indeed still absorbed in the ground. The issue of whether or not it is practical to harvest that energy is one thing, but capturing that energy that is being produced and dissipated in any case is at least possible without increasing the work done. One can even imagine a system located on a set of down only stairs in which the human work of movement is primarily eccentric (slowing down the lengthening of muscle by triggering partial contractions) - in such a case the system harvesting some of the energy of descent means that small amount less eccentric work the human needs to do. (Again, the parallel to regen braking in which the regen helps decrease how much the brakes need to do and how much heat is produced by the brake pads.)
Think also of a pedestrian bridge that vibrates some as many people walk across, vibration that the designers would like to dampen. I’m no civil engineer but I imagine that one approach to that is passive dampening that ends up releasing the energy as heat, and that same energy could, theoretically, be reclaimed for electricity.
Even in the exercising human scenario. Hard to imagine how to do it but the point is that there is heat lost by the human exerciser, the human in fact needs to get rid of that heat. Some numbers: during “180 s of intense dynamic knee-extensor exercise (≈80 W)” young volunteers were needing to release 24 ± 1 J s−1 at 180 s from the knee extensor muscles alone, having increased during the exercise period. Now that generated heat is generally lost to the environment, via radiation, convection, evaporation, etc. such that body temperature is maintained within an acceptable window, but theoretically some portion of it could be harvested and converted without increasing the work being done by the human.
The overarching point is that of the energy used by human muscles only a fairly small potion of it is used to move us, call it “work against gravity”. Some is conducted to the ground, some is used deforming our clothing, and a fair amount is heat that is necessarily transferred into the environment. Theoretically some of that energy that is being expended in any case could be harvested.
Of course a human doing a bit more work during the day is a potential selling point not a downside: “Powerwear™ -not only does it power your device but you can burn 300 Calories extra every day! Without any extra exercise!”
If a human is deliberately exercising for the sake of exercise, then it’s easy, because that lets you change the parameters of the exercise to something that’s easily harvested, like riding a stationary bike.
Yeah, that bit’s been done … just more expensive than worth it other than for greenwash marketing points. It is not difficult; it just aint cheap.
Probably the electronics that react to movement in clothing will be more used to provide data for fitness tracking and health parameters than for power, other than possibly powering the clothing embedded sensors, transmitters, and such. It really is not too far of a … stretch … to imagine an UnderArmor style compression outfit that sells telling the consumer more precise activity (including non-exercise activity and data like vagal tone) information than any FitBit is capable of, and is self-powering. Or running shoes that report not only how many strides you have taken and the length of your stride but report bak the exact nature of your strike.
The next step would be materials that have tunable characteristics in response to the data collected. (For example reflecting or conducting more or less body heat.)
I just saw a segment of a show on PBS last week about a team of high school kids who developed their own power generating floor tile. The went from bulky wildly impractical proof-of-concept model (looked to be about as thick as the ones shown in the video you linked) all the way to working model refined enough for production. They were trying to power ultraviolet (I think) water purifiers for use in Africa.
(in a quick google search and search of the PBS website, I was unable to locate the show to link to it)
The trick with this type of device is going to be avoiding modifying existing building or areas just for the sake of this thing. Primarily new construction, or as part of a planned maintenance program, is where this will be best applied. In this case, quantity is what you need for meaningful power harvesting.
You’re never going to get this sort of thing to be economical. Human beings suck as a power source. The only real application for it is in powering small portable devices that can’t be connected to an electrical socket reliably. So things like watches, phones, hearing aids, insulin pumps, and so on. The other possibility is for remote areas that aren’t connected to the power grid. But in that case you’d want ways to generate lots of power, rather than a few stray watts, and so you’re looking at things more like a stationary bike. But you might as well set up a windmill or solar panel by that point.
Or, as Chronos pointed out, using human muscles to move human beings around. Take the stairs instead of the elevator, ride a bike instead of take the car, fan yourself with a hand fan instead of sitting in front of an electric fan, play the piano instead of listening to the stereo, get up and walk to the fridge for a cold one instead of ordering your robot butler to get it for you.
Why do you hate America? 
Wow, thanks; this is what I was thinking about when I asked the question.