so, here is some data:
the fluctuations you see is from the temperature changes in the room. i had to set the AC and heating about 8 degrees of space between them to prevent them working against each other and not kill the electric bill of my shop.
as this is a wear test and not a lab test so this does not invalidate the test unless you drive around in perfect 20c weather all day. in that case i want to know where you live and what the housing market is near you because i want to live there.
tests done with brand new samsung E29E7's.
benchmark cell: 4.2V~2.5V with 1.38A charge and 2.7A discharge:
note that i stopped at 450 cycles as i am moving my setup to a different workshop.
https://i.imgur.com/JChi55E.png
next up: the same but with 4.1~3V 500 cycles test:
https://i.imgur.com/D7fZlp8.png
next 500 cycles at the limit of 2.7A charge and 8.25A discharge at 4.1V and 3V:
https://i.imgur.com/mTLMZ8r.png
note that the high current tests are more even because they heat up a lot more.
so in the end the conclusion is pretty simple: as long as you keep under 4.1V and above 3V you can pretty much do whatever you want, cycling beyond those numbers does WAY more damage then running them hard does.
but as seen in the second test, if you baby them then they are pretty much indistructable.
this is bascially also how tesla gets away with their battery warranty, you need at least 1500 cycles or more to make a noticable dent in the capacity as they have way more concervative voltages (4.05 and 3.1V).
fun fact: when running the full range capacity thest the cells that had their full range used (the benchmark cell and the last high current cell) immedialty tanked to 3.8V under 0.8A load while the other cells only dropped to 4.05V.