Nifty article. It reminds me of the books that Oliver sacks wrote in "An anthropologist on mars" and "The man who mistook his wife for a hat" on some of the weird things that the brain can throw up if it gets damaged in a particular way.

As for scratching through the skull, I can see how it would be easy to use your fingernails to wear a hole through the skull given enough time and patience. Fingernails grow a lot faster than bone can repair itself, plus the constant state of infection an inflammation would help in weakening and breaking down the bone itself.

Nanuq, I think you have a winner! from the great Wikipedia, Gold has a density of 19.3  g·cm−3 and tungsten has a density of 19.25  g·cm−3. the bar would be lighter enough to be detected easily by the Archimedes test. They also have pretty different thermal expansion coefficents, you might be able to force the gold to de-laminate from the tungsten interior by dropping it in liquid nitrogen

I can't imagine why not. the tricky part is concealing the mine/sub from an active ping. A ping is a close approximation to an impulse function, which is effectively an infinite series of discrete frequencies. To make a "perfect" cloak, you would need to be able to effect enough of these frequencies to fool the sensors of your opponent.

Though talking about the mines vs. sub, I have no idea how radiation within a meta-material volume would react, which is critical for an active system like a submarine. Would it reflect the acoustic energy back inward, shaking the sub apart? Or would the energy be refracted outward in some strange way that would be easily detectable? There is probably a PhD thesis in there somewhere.

All I can say, is if they manage to make a wide bandwidth material, absolutely brilliant.

[Warning, High Science content ahead!]

The idea is that you can use all of the nifty tricks that the Electrical and Optical engineers have been messing around with in making negative index metamaterials. The brillant bit is that in this application you are treating light as a wave, so why not use the same math for a different type of wave like a sound wave? You basically have a 3-d array of resonant units, so I imagine that an acoustic meta-material would resemble some sort of structure with a bunch of small tuning forks.

The current problem with negative index meta-materials, is that the "bandwith", or number of different frequencies, is pretty limited; also most meta-materials are pretty lossy. This means you might be able to make a column invisible to high notes or low notes, but not both. Still, this is really only an engineering gripe right now, there is nothing to say it won't be solved.