The electromagnetic effect falls off exponentially the further the coil is from the metal to be heated. In the case of a large wrap, you want it as close as possible.
I did the math to calculate the energy to melt the metal with the lowest melting point in the system. In the time to boil one gallon of water, it would take 94kW to melt a similar volume of aluminum. Your residential power supply isn't going to be able to put that out, so you are probably safe. I assumed 1 gallon of aluminum is what you are going to melt, based on the fact that aluminum is decent at distributing heat and would hopefully be cooled by the water. However, if it was thin walled, on its own, and not redistributing the heat, or had impurities had was highly resistive to distributing the heat, you could melt it with as little as 9kW... at which point you could have a serious problem. Of course, your metals are probably alloys, rather than pure, so there is a lot of variance for melting point, slag point, and so on.
However, if there are any impurities in any of the metals (which is why I used the term slag), you may have some burn off.
I have two worries with any induction system: 1) electricity, 2) distribution of heat. First, you are running electricity through your coil. You will have to insulate it. If it is not completely insulated and it comes in contact with your pot - you will die if you ground it out - whether via a hand in the water or the keggle itself. The insulation will have to be able to take the heat. However, the electromagnetic force of your coil reduces exponentially with distance, so you need to keep it fairly close to what you are trying to heat. Second, a coil will heat things inside and outside of it. If you are wearing a watch, have sensitive electronics, or have a wedding ring with any content that is magnetic, it could heat up. Furthermore, if there are any air gaps between what you are heating and the water it contacts, it will heat up quickly and increase the chance of fire, burns, slag, and other problems. Slag is a problem if there are any impurities in the metal. Those impurities could burn or melt off.
Furthermore, the diameter of the coil is going to change everything also.
This doesn't mean it cannot be done. It means you are going to have to look at existing designs and work from there.
If I was going to go the route you are suggesting (which I have considered in the past), I would go with a stainless steel keggle that has a high iron content. The only way to test that would be to take a magnet with you and see if it sticks. If it sticks solidly, you are in business.
By putting the keg in the center, you are building a coreless induction furance. You would have to have heavy duty, high conductivity copper tubing wound into a helical coil. That coil would have to be wrapped in a steel shell (double walled stainless steel keggle would be great for this, as you could put it between both walls of the keggle). On the outside of the coil, you would have to put magnetic shielding to prevent heating of the supporting shell and to protect anything outside from being heated or experiencing electromagnetic interference. You would then have to pump a coolant through the coil to keep the electrical resistance down and to keep it from melting. Most likely coolant would be water. If you wanted to be efficient, you could use either water going into a reservoir for cleaning (that would be a ton of water in the end), or water going into a cooling tower (radiator) that releases the heat and comes back in. I would consider doing it as two steps, the first step would be circulating the wort through to use the outside heat to heat it up, then once it hits a set temperature I would run water in to clear the line and then circulate water through a radiator to cool the unit. The nice thing here is once you turn off the coil, you can keep the water running and use that as your wort chiller.
Now, you are going to have a problem if you ever fill the keggle with less liquid than the max height of the coil, because the metal not touching the water is going to get super hot.
Of course, this is just the coil! Now you have to build the array for increasing the frequency, maximizing the timing to voltage ratio, settling your curves so they're squared off, and keeping the capacitor unit cooled.
This system would be very bad ass, especially if you could set it up to run on a 120. However, it would be significantly easier if you had a 240 or 480 tri-phase or quad-phase power supply lying around to be used.
And that is why using a cored induction coil in a heat pump system would likely be easier, not to mention it would definitely be more efficient.
That help?