I think the way to "balance" the use of free energy may very well come down to this:
First, you can reuse energy over and over. The only real conservation of energy law is that "energy can neither be created nor destroyed. Period."
This means that, when you "use" a joule of energy (e.g., suppose you scatter a joule of incoming energy, in a resistor), then you still have that joule of energy remaining AFTER you've done a joule of work.
So in May this year at the 4th International Energy Conference, I pointed out that you can RETROREFLECT the energy, after it's already "passed through" the circuit, and get it back (or some of it, depending on the reflection coefficient) to "reuse" (i.e. scatter) once again.
I pointed out that this already occurs in intensely scattering media, and provides the well-known anti-Stokes emission phenomenon, which IS DEFINED as emission by the medium of more energy than one puts in! In other words, the fact that one can get overunity has been recognized in the anti-Stokes emission phenomenon for over 30 years.
Now all energy comes from the vacuum. Sorry about that to the electrodynamicists, but if they will read a little particle physics, it has been known for 40 years that any electrical charge or electrical dipole is a "broken symmetry" in the violent energy flux of the vacuum. That's a proven thing.
By definition of "broken symmetry" in a virtual particle flux (which is what the vacuum energy is), particle physics tells us that
(1) that asymmetrical portion (i.e., the "gated" portion or "extracted" portion) changes from virtual to "observable", and
(2) that observable energy flow component will be radiating away from the charge or dipole.
Note that electrodynamics doesn't even know what the whack an electrical charge is, much less a dipole! There is presently no definition of electrical charge in physics.
(1) already freely extracts energy from the seething vacuum, due to the dipole's asymmetry in that flux, and
(2) already puts out an observable flow of energy as a result.
Well, our electrodynamics profs forget to tell us that! Because it doesn't appear in the 130-odd years-old electrodynamics model! But it's been known for over 40 years in particle physics. It would be nice if the various disciplines would "scrub" their older stuff and correct it, when later and different information uncovered in science has falsified it. They don't, at least not in electrodynamics.
Anyway, we have these facts from particle physics:
(1) an electrical charge freely extracts energy from the vacuum, and
(2) it outputs that energy as a radial flow away from it, in all directions, and
(3) that energy flow is of the kind that we call "observable" (i.e., detectable in that it will translate matter under the right conditions).
Well, there is only one kind of observable energy flow from an electrical charge! And that is the so-called Poynting flow S. (Simplest form is S = E x H, although there are expressions also for the absence of the magnetic field and so on.
Also, the real thing is S = E x H + G, as pointed out by Heaviside who independently discovered it about the same time as Poynting. Poynting got the direction wrong by 90 degrees, and missed the other factor G.
In fact, Poynting effectively limited the energy flow vector S to the energy collection flow in a circuit (the Slepian vector j-phi, for energy density flow).
Well, that throws away almost all of the energy flow extracted from the vacuum by a source dipole in an electrical circuit. Let's put it this way.
The source dipole extracts freely from the vacuum an enormous flow of energy, and sends it out along the external part of the conductors in its external circuit. All space around those conductors (transmission lines) is filled with this Poynting flow.
The total energy flow is enormous. But only a tiny, tiny smidgeon is intercepted by the surface charges in the wires (and components). Just that tiny portion of S that is in a small sheath directly adjacent to the conductor/component surfaces, gets intercepted and diverted (diverged) into the circuit to drive the electrons as j-phi.
And thereby hangs a tail. All that Poynting got in his derivation was effectively that divergent part. He also got the direction wrong. Heaviside corrected him as to the direction, and also pointed out that THE DIVERGENT COMPONENT OF THE POYNTING FLOW IS ONLY A TINY PORTION OF THE ENTIRE FLOW.
Most of the flow in the space surrounding the circuitry DOES NOT contact those surface charges, DOES not get diverged. So the nondivergent component is huge (nominally about 10exp13 times as much energy flow as our feeble circuits intercept and collect from the small, small local sheath in contact). The divergent component is about 10exp(-13) of the entire S-flow.
Now contrary to the textbook, the Poynting flow is still in virtual form, but it is in ORGANIZED virtual form. So our normal circuits have
(1) taken enormous energy from the local vacuum, and
(2) put back almost all that energy, in flowing nondivergent form.
The reason our feeble electrical circuits do not produce lots of general relativity phenomena, stopping watches and clocks, time dilation, etc. is because they are so damned inefficient, almost to the point of extinction!
So you can say that the normal inefficient circuitry only has a "general relativity reaction coefficient" of about 10exp(-13). That means you normally don't have to bother about time effects, curvature of spacetime, conditioning the vacuum, and oddball ST curvature effects on minds and biological systems. The vacuum doesn't locally "condition" very much.
Now if you were to materially increase the efficiency of that localized "interception and divergence" (i.e., collection) by the circuitry or medium or whatever, then as the orders of magnitude increase, you can get appreciable gravitational effects and all the other. In other words, now you are taking out a lot of local energy as virtual flow, BY PROCESSING IT THROUGH THE CIRCUIT AND ITS LOAD TO FORM "OBSERVABLE" PHOTONS.
The precise difference is this: Putting back virtual photon stuff immediately restores the ether (vacuum). Putting back observable photon stuff is actually putting back DISTURBANCES in the overall magnitude of the vacuum, NOT local virtual photon content. And thereby lies the rub!
It means if we get a good "free energy" system, but do work that produces almost totally OBSERVABLE scattered photons, then we have a local vacuum problem.
We are drawing out the energy in its native form, but we are not putting it back in its native form. So we are "continuing to enlarge the hole" in the natural state energy of the local vacuum. And that gives us the problems.
Well, there is an obvious technical solution. We must REPROCESS those observable photons scattered by our loads, back into virtual and uncorrelated photons.
In that case, we would be back to our present power situation: We could draw, collect, and utilize megawatts or gigawatts of power (easily, if we accent those principles of increasing the collection in the circuitry or medium, as I gave in May 1997) if we are careful to reprocess most of the scattered "waste" photons from our work processes, back into virtual photons which inject directly into the native vacuum format.
There's another way to help also. When we draw out a certain rate of NATIVE energy format, and process it into observable format, then instead of just propagating that waste observable energy out of there, we can RETROREFLECT it repeatedly and intensely, and reuse it.
That way, we dig a certain size "hole" in the local vacuum native energy, so to speak, and then stop. We reprocess most of the waste energy, so our overall POTENTIAL (i.e., density of collected energy) can be very great, without getting into all the weird "vacuum conditioning" effects.
That is, we store up an intense reservoir of recycling observable energy, and use and reuse that energy over and over. In the real world, we can only do that so much, but in an intensely scattering medium that can be appreciable (as shown by Patterson's macroscopic adaptation of anti-Stokes emission and by Lawandy's use of colloidal solutions of titanium dioxide (main ingredient in white paint) to achieve great energy gain and lasing without population inversion).
In other words, you set up a given "storage" of "cumulation" of recycling energy. The actual linear flow of the energy can then be enormous, without taking out nearly that much from the native vacuum energy format.
You only have to replenish the losses or inefficiencies of the recovery process.
Say, e.g., that you can get good efficiencies in cumulation recovery so that only 10exp(-6) of the energy is being irretrievably lost. Then you can power substantially large loads without too much fuss from the native vacuum.
AND, you can also apply to those losses a "reprocessing into native format" process, to further reduce the local vacuum environment's degradation.
I have wished for some time that we could get the Department of Energy and the Universities turned on to this type of thing and to overunity. That still seems to be almost a lost cause. They simply can't seem to get it through their heads that only by altering Maxwell's equations (regauging symmetrically, the Lorentz condition) have they ASSUMED AWAY the overunity systems that Maxwell's unregauged equations rigorously allow and prescribe.
And they can't seem to get it through their heads that (electromagnetically) the work-energy theorem W = (phi)(q) is a special case for ONE-PASS of the Poynting energy flow along and around a circuit, and on off into space.
(You get a completely different version when you use multiple passes by retroreflection (particularly by phase conjugate reflection).
Then you have to use something like W = (alpha)(phi)(q), where alpha is a factor due to multipass, multicollection. In that case, alpha <1.0 gives you the conventional case, alpha = 1.0 gives the limit of the conventional case, and alpha1.0 gives you the multiple pass, multiple collections case as demonstrated by anti-Stokes emission.
Anyhow, I hope this overly long expose' helps clarify the problem of powerful overunity mechanisms initiating dramatic alterations of the local vacuum and of spacetime, unless certain precautionary measures are taken.