In Dave's video, towards the end, the battery voltage was far too high, for a run down battery (approx 1.25V at about 1800 Sec into test). I.e. I smell a rat. (Tomfoolery/trickery).
No trickery, the electronic load had switched off because the battery voltage dropped to 0.1V. What you saw was the battery voltage recovering under no load.
The voltage seemed/seems too high, to recover that much, from being forced down to 0.1V, for a while.
But you could be right and/or maybe Dave took a break or something, giving it time to recover.
Also the second battery (B ?), had much more time to recover (maybe not, I could be getting confused, because of the sudden time jumps in the video), than the first (A ?). That could also be the main factor in why we got the results shown.
Really, both experiments, should occur at exactly the same time, to be fair/accurate. (If you only have one of each of the test items, it is harder to do it at the same time, But still can be (done, by using the wall clock, and doing 2 passes of the experiments), by recording the physical time, and keeping the relative times, about the same, between both batteries. I was not sure, if that was or was not done, for this experiment).
BUT I could be misunderstanding the timings, from watching the video. Maybe the timings were NOT that far apart. It is NOT very clear, what the overall timings were (to me, at least).