Interesting articles! But as usual there's no free lunch. The aluminum in the first example needs to be "recharged" or refined back to its initial state which requires an energy input. That puts it in the energy storage medium class (battery in my way of thinking), so the energy input to return all the various parts to the original state needs to be examined and rated for its "energy to the wheels" efficiency.

The other two appear to be the same process, which as was noted is about the same as adding Draino to aluminum cans to release hydrogen while oxidizing the aluminum.

Electrolysis is a reversing of the hydrogen-oxidizing process and because of the inefficiencies involved will be a net loss. As usual, the entire process ending with energy to the wheels needs to be looked at. If electrolysis was 70% efficient and the fuel cell was able to release 75% of the remaining energy as electricity (both numbers on the optimistic side), the overall efficiency to the wheels would be well within the range of the present internal combustion engine because you still have to account for the energy production for electrolysis.

I'm not against it; part of one of my jobs in research and development was solid oxide fuel cells so I know the advantages. I also know the disadvantages.