If that is the goal, then write an article about it.
See above - I already posted a short exposition by an expert in the field.
Again: This article is about when batteries run empty. The voltage recovery in this case is more than twice the value compare to a battery with some energy left and if it happens instantaneous or over some time do not really matter
For
qualitative understanding it matters little that the bounce is bigger at the end of discharge. What matters is that it is significant and prevents said naive measurements from succeeding. The same holds true for the termination of any (dis)charge at any point, or, more generally, for any nontrivial change in current (for nontrivial time). But this will not be clear from what little is said in the OP.
(The model your published covers neither case).
The models convey accurate
qualitative understanding of the reason for the voltage bounce. Precise
quantitative understanding is not required for the above purposes (nor is it conveyed
generally by the single graph in the OP).
The interesting part is that there is a significant voltage recovery, it is fast enough to prevent measure the discharge end voltage and it varies with a couple of factors.
Yes, but simply pointing at a graph and saying "this is what happens" is not the same as imparting some conceptual understanding of the underlying phenomena. Unlike a graph, models apply generally and allow one to make very general inferences. For example, from what little is said in the OP it will not be clear that, by symmetry, a voltage drop occurs at the end of a charge, and that these bounces are (roughly) proportional to the current, and that analogous bounces occur with any nontrivial change in current. These and related inferences are all immediate if one understands any basic model incorporating internal resistance (which requires only basic knowledge, viz. Ohm's law).
Of course one needs more complex models to deduce more precise
quantitative information - which requires understanding how internal resistance depends on various parameters (SOC, temperature, etc), and the role played by time constants of various internal processes, etc. But none of this complexity is required in the OP.