Ultracapacitors will never reach the power density of batteries. The main reason is that the battery can lay down thousands of layers of a chemically changed state at the change of state voltage.
The capacitor can lay down a single layer, at first, when the second layer is built, a space charge analog effect occurs, sine the 2 layers have the same charge polarity and repel each other.
Capacitors are made to have a very large area, far larger than batteries, and they can discharge this rapidly and can be repeatedly charged and discharged, since there is no wear mechanism. With capacitors, the charge is a function of voltage, as you draw charge, the capacitor voltage drops. Batteries are flat in voltage profile until the battery substance has deplated,m it then falls to zero.
The mechanical analog is a spring, which flexes and the force declines, the hydraulic analog is a dam -which empties

Obviously, if a teen finds a bug, he should offer it Romanian hackers, cash first, in an anonymous way.

The analogy equates the voltage to a water level to an absolute temperature.
Sure yoiu can use voltage regulators in boost mode, but then you could use a discharge temperature of absolute zero to get more from a Carnot limitation.
( turns on the port to his infinite volume absolute zero discharge space - which he carries on his back...)

Another limitation in capacitors is encountered when you increase the voltage - this means thicker insulation = lower capacitance in inverse proportion.
Double the insulation = twice as big in size. The stored energy goes up as the square of the voltage, so there is gain with voltage.

The big limitation is that charge is only a single layer on the electrolyte, so you can never achieve the storage density of chemical change of state, nor the constancy of voltage. Added layers do not allow for direct increase in capacitance. A full double layer will give an incremental increase. Batteries can form layer after layer of stored atoms in the charged state with directly additive results

One day they may solve the dendrite problem in silver cells - as they are recharged the silver forms atom sized pointy needles that work there way through the porous plate barriers (they must be porous to allow current flow). So far this has proved intractable.

No, capacitors will never replace batteries, they will complement them.
The charts and work of others are valid.

The voltage declines as you consume the energy stored, meaning some of the energy you stored you can never use.
This is analogous to the Carnot limitations.

I took this in 1958 and I was unable to cut and paste from the Wikipedia.

That depature from reversibility indicates that Li-ion will inherently lose part of their charge

http://en.wikipedia.org/wiki/Capacitance
http://en.wikipedia.org/wiki/Capacitor

Yes, I took this in 1958, so I am a little rusty on the terminilogy, bit clear on the concept

One problem with capacitors is the charge is stored a lot like water in a tank. As you use water the water level drops, in any capacitor, as you use it the voltage drops.
The governing equation is Q = 0.5 *C*V*V.

A single cell (in a battery of cells) is composed of two materials of different chemical states and they produce a constant voltage until one of the chemical states is depleted. Charging reverses this, again at a constant voltage. The charge and discharge voltages in a theoretically perfect cell are ~~ the same, in a real cell, resistance caused voltage drops and departures from irreversibility lead to differences in the charge discharge voltage. You must charge with a high voltage than you get on discharge.

A second problem, is the fact that a bulk material changes state in a cell, this inherently stores more charge than a capacitor, which is a surface layer of added charge. It is true that since the capacitor involves no change of state, that the life is more or less infinite, and because it is a monolayer of charge, you can charge and discharge at speeds limited only by the current limits of the wires.

The net result is the energy density of the best capacitors is barely as good as the worst batteries.
Battery graphs here http://tinyurl.com/autjb7l
Capacitor graphs here http://tinyurl.com/byqbdje
Direct comparisons here http://tinyurl.com/b9zwcdw

As long as you design a downstream voltage regulator to use the declining voltage to power your circuit at its required constant voltage, then ultracaps will find a niche in many pieces of equipment from Cars(as a peak acceleration source) to tiny items as the sole power

Robot workers will get paid and transfer their earnings to robot consumers, a closed cycle capable of great speed.

If in fact they automate all work, and all people do nothing in the way of work, then output will soon fill the warehouses and people will buy nothing - they have no money to buy anything with.

Or will we operate like a star ship, with no money, all done by "synthesizer rations"

Let states charge their own sales tax on sales that start and finish in that same state. If they ship to another state, charge 7% Federal interstate sales tax and keep 33.3% of that for the federal government, and give 33.3% to each state. States would get some of what they wanted, and on both ends and the feds would get their third.
Once passed, states could opt in or not. If they do not, they face what they face now on sales shipped into the state?)and they get nothing on sales shipped out of the state. My bet is they will take the 33.3% into and the 33.3% sales out of as better than no loaf.

Yes, current 3D printing is a gimmick. I want to see 3D printing with embedded metal parts and semis, circuits. Now all you can do is make a hollow that a metal part of board can be inserted into - ideally snapped into with a positive lock, so you can add wheels with bearings. Circuit boards, LEDs etc can also as snap ins.
A software library for these snap ins is needed, with dimensional adjust-ability. What is also needed is deductive post 3D printing, -sandpapering irregular bits, and also the ability to machine it after the first print run - insert the bearings and circuit bits, a 3D print some more.
As it sits, it is barely OK for small runs on some stuff. Injection molding is far better for volume.

One day, they will make these with 30-40 bins for various elements and you will be printing at the atomic level and you can make an iPhone... LOL, see this space in 20 years and it may have come true?

12.4 cents/KWH here in Canada. The path to recycle this aluminium is more tortuous and costly, but with enough cars buying new AL battery inserts and dropping off oxidized slush they will find a way, even if not economic compared with new mined alumina where the power is. If this works it will take 40-50 years to be put in place wlong with the others, like Vanadium and Lithium.

We need to assess the overall efficiency of the process that uses aluminum in the battery and then electro-refines it via the Hall process.

No, It is a good idea. It would get tax to both states and the feds abd be a great leveller of I/S commerce

The feds should impose an interstate commerce tax, say 9% and give 3% to the ship-from state and 3% to the ship-to state and the feds grab 3%.
States with no sales taxes, their 3% is omitted.

This will give states a bite of in and out traffic, that they get little of now. Not as much as the states full taxes, but they lose most of that now. 3% of both ways is a lot btter than what they have now. It will give the feds something to erase debt, it waill act as a leveller of the playing field.
States will have to waive their state use taxes on good shipped to to conform with this law.

As we gradually change from ipv4 to ipv6 pressures will emerge that cause migration to ipv6. As the world as seen by an ipv4 gradually shrinks, the users will clamor for ipv6. Some dead ends might stay with ipv4? As the current low (~1-2%) rate of ipv6 adoption, the ipv4 group do not suffer greatly, but as allocatable blocks dry up, and the cost of ipv6 hardware declines there will come a tipping point which will place inexorable pressure on the ipv4 to switch instead of using translators or tunneling methods, which bottleneck them even more.

On cellular networks there are hard wired protocols that all must adhere to. recall the transition from the analog to digital methods a few years ago? A sunset time for analog was set and after that date(which varied by country) the analog went dark. Similarly, all makers of gear will have to adhere to a group standard to enable operation at all. I do not know enough about the tech to know if high speed cellular 'converters' will allow mixed operation? One suspects the overhead will be oppressive.