I take a very dim view of any manufacturer that believes series charging lithium cells is ok, with them falsely trusting that natural balancing is ok. It is simply not plausible the pack manufacturers are grading cells for capacity and internal resistance that are within 0.5% of each other before assembly.
Speaking of which, Sony’s chemistry, now owned by Murata, has the best consistency and after 500 plus cycles will still match within themselves the best. Often batteries built with competing brands cells, when pulled apart at the end of their service life, one cell is letting the battery down. And this is in a fully balanced pack with a proper BMS. Your unmanaged, series changed battery isn’t making it to 500 or more cycles. Even with a BMS, you cannot just replace that now faulty cell and return the pack to service. What goes in must match it’s neighbours for capacity and internal resistance.
A murata VTC6 is hard to go past when lifecycle is considered, but…..
Things significantly enter the region of “currently unknowable” when we introduce MTB vibrations into the equation. For road going EV usage, testing has revealed 18650 lithium-ion cells degradation is predominantly charge/discharge cycling, that vibration is an order of magnitude smaller with regards to cell capacity degradation.
I’m planning on a test jig, that has 2 of each cells of interest mounted to a vibrating platform that moves in X y a roll and yaw planes to discharge while being vibrated, and charged while stationary, three times a day hopefully to get some data on degradation in a mountain bike scenario.