We may see an arms race between the energy harvesters and the energy wasters. All the energy that gets harvested (and it is a small fraction of the total ambient wasted energy) is being produced by something. If the energy producer better balances their motor, they reduce their wasted mechanical energy (while cutting off the harvester). If an electrical system is broadcasting unused energy, it is being inefficient. I'll be interested to see how systems start reducing radiated energy or recovering the energy to improve efficiency rather than just leaving it there for a few random devices to harvest.
In the most extreme cases, I wonder if the energy harvesters will start putting a measurable load on the energy producer and forcing them to increase their energy consumption to cover their work as well as the parasitic loads.
The current Imec/Omron prototype generates 70 microAmps with at 1 volt would be 70 microWatts--10 for the sensor and 60 for the comm--plus they claim to be working on upping the output. However, its not a product yet00-only time will tell.
Well, to read at 10 uA, 1V, there is already 10 uW involved, and obviously we need many times more than this power for processing, so I do think energy harvesting is currently insufficient to meet energy needs. And don't think energy harvesting doesn't have its environmental consequences. Energy that is harvested comes from somewhere, and we expect that energy will not be replenished at the source?
Most of the traffic IOT (IOT that is outside relaying traffic, both human and vehicular, data to clouds) will benefit from this kind of an energy harvesting technology because IOT power generation is a hotly debated topic, and this seems to cool down the waves of people saying IOT cannot be funded because the energy costs are too high.
Imec/Omron seem to think there are applications in that range. Here is what they said when I sent them your comment: "The energy harvester can generate up to 70 microwatt of DC power at the required (user-defined) voltage. It starts generating power from very low vibration conditions ( < 0.1 g). To operate a battery-less wireless sensor node 5 to 10 microwatts is sufficient, so even at low vibrations this power supply offers a solution for low-power wireless sensors."