"easy to come by" can be quite misleading.
https://en.wikipedia.org/wiki/Prices_of_elements_and_their_compounds
I hope nobody will protest the use of wiki prices?
if I/we assume economically abundant = under $10/kg, list of economically abundant under $10 elements
Zn, Si, Sa, O2, N2, Mn, Pb, La, Fe, H2, Cu, Ce, Cl, Cd, As, Ar, Sb, Al (the 2 most expensive of the $10 bracket are H2 (liquid) and Cu).
sulphur ($500), sodium ($250), at $250 (+/- $50) also sits our famous friend lithium and tellurium.
I thought that carbon should be really cheap, but carbon is $24/kg, similar as Sn.
Sulfur is incredibly cheap; that might be the price per ton(ne).
Check the refs. A lot of those look to be reagent prices, which bear no connection to reality.
Carbon might be that expensive if very pure (possibly... nuclear grade graphite?), but metallurgical grades (such as coke) are in the same bracket as iron.
Similarly, sodium in large quantity is limited by the same economics as aluminum, magnesium and such: it is metallic energy. I imagine the bulk price remains higher than those, due to smaller production quantities and more hazards to production and transportation.
Funny thing about arsenic, there's a tremendous amount of it available in various areas; supply greatly exceeds demand. For obvious reasons. It should give higher voltage in a battery -- but I suspect the cost of using it safely, greatly surpasses its usefulness in the end.
The case is made of steel, or stainless steel, so it doesn't melt.
The issues to solve are probably more those of amalgaming, contamination, and secondary unwanted reactions among all the materials present.
Mg has absolutely no effect on Fe. In fact, the only connections I know of between the two are both in foundry work:
1. steel containers are used to handle molten magnesium (exactly because it doesn't corrode iron).
2. small additions of magnesium, to cast iron, encourages nodular instead of flaky graphite crystals, making ductile instead of gray iron. (Graphite has essentially no tensile strength, so flaky graphite crystals embedded in the iron matrix basically act as stress raisers -- myriad microscopic cracks. Gray iron has no ductility, malleability, and poor tensile strength. Ductile iron, as the name suggests, improves all of these properties. Malleable iron, is gray that's been heat treated, producing a similar morphology at the expense of several days' soak at orange-hot temperature.)
What I don't know about, is antimony's effect on steel. Does it passivate? Could stress corrosion cracking be a problem?
The most important I think, as hinted in the talk, is the insulator. Most ceramics have permeability for various ions, at elevated temperature. Beta Al2O3 for example is used as the sodium ion-bearing electrolyte in NaS batteries. ZrO2 has oxygen ion mobility, hence its application as a combustion sensor in vehicles ("O2 sensor", but there's no O2 in the exhaust stream of course..). I don't know offhand what they'd be looking at. It also needs to have close thermal expansion to the metals it's swaged into.
The terminals could very well end up a dominant part of the cost, needing smaller amounts of much more expensive materials.
Tim