Energy density has a fixed upper bound for any given battery chemistry: the amount of energy that would be bound/released if all of the material in the battery would participate in a cycle by turning from one state to the other and back. The amount of energy per reacting molecule/mol/kilogram is fixed and well known. Someone who is not as terrible at chemistry as me won't find or hard to come up with concrete numbers. And those numbers are hard constants, you can only come closer to them (within a given battery chemistry), they are impossible to exceed.

And another angle: our perception of energy density increase is heavily disturbed by the much faster reduction in cost, we are just not very good at telling those two apart.

true but how far are we from chemical theoretic optimum ? manufacturing of cells is very suboptimal at the atomic level.

The trouble is making the cell chemically stable with a way to get current in and out. We can't know what configurations we'll find where this works. Maybe tomorrow a catalyst that changes everything will be found and our batteries suddenly have the energy density of TNT.

Permanent magnet strength also follows a similar pattern, the strength doubling roughly every decade or two.

Even more interesting

I'll call it Rick's law: Roughly every 4-5 years, Li-ion battery prices will halve

Let's see if it sticks