Per each electron provided in the circuit, the mass of aluminum is 9/7 of lithium and the maximum voltage is around 93% of that of lithium, which results in an energy per mass for aluminum of around 73% of that of lithium, when only the mass of the metal is considered.

In an aluminum-air battery vs. a lithium-air battery, the mass per electron is, in the most favorable case for lithium, of 17 for aluminum vs. 15 for lithium, which results in an energy per mass for aluminum of around 82% of that of lithium. However lithium forms by oxidation not only Li2O, but also peroxide Li2O2 and superoxide LiO2, which may worsen a lot the energy per mass.

In the parent article, they have succeeded to produce mostly Li2O, but even so their batteries have still produced some amounts of peroxide and superoxide during deep discharges.

So the energy per mass for aluminum-air batteries could be up to 80% to 85% of that of lithium-air batteries.

Most other oxidants besides the oxygen from air are heavier, which would reduce the advantage of lithium vs. aluminum (because the oxidant mass would be a greater fraction of the battery mass), so aluminum-ion batteries, if possible, could have an energy per mass very close to that of lithium-ion batteries.

On the other hand, aluminum metal and aluminum oxides are much denser than lithium metal and lithium oxides, so aluminum batteries could have much better energy per volume than lithium batteries. Unfortunately, until now the problems caused by aluminum as a cathode material have not been solved.

Alcoa and Phinergy demonstrated an aluminium-air battery over a decade ago that enabled 1600km range. The downside is that it had to be swapped out and recycled after it was spent (which explains why Alcoa's interest since it would gain business from the ongoing re-processing of aluminium).

https://eepower.com/news/ev-with-1000-mile-range-unveiled-by... https://www.cbc.ca/news/science/electric-car-with-massive-ra...