Theoretically this is what the manufacturers do
I think you are somewhat missing what
@snoobler and
@Gazoo are trying to convey.
I believe the point both are tying to make is not that no loss of capacity will occur, but that capacity loss over time discharging to 2.5 is within the parameters that the manufacturer considers normal/acceptable. I.E. the cycle life count on the datasheet (2.5k loose, 3.5k with compression) is based on discharging to 2.5. So by saying you are not 'harming' your cells discharging that deep it is not to say you are not shortening their life, but that it is within acceptable (to the manufacturer) parameters and isn't causing substantial or significant harm if done occasionally. This can be true, while at the same time your point can also be true (a higher LVD / tighter charge-discharge bandwidth) will increase the cycle life of the cells). They do not contradict each other necessarily. I suppose to some degree it comes down to how you define 'damage'
I believe the only point that either are trying to make is that the occasional deep discharge won't significantly and immediately damage the cells and that manufacturer cycle life claims are (often but not always) based on full 2.5-3.65 cycles.
Some caveats and points to ponder:
- The EVE datasheet seems to extrapolate cycle life from ~600 test cycles, I am neither a chemist or statistician so can't say how representative this is of the full cycle life.
- Calendar aging and temperature are wildcards that may greatly affect cycle life, and neither are properly reflected by manufacturer test conditions.
- There is a decent bit of variance in claimed cycle life as well as test conditions manufacturer to manufacturer, EVE is at the moderately optimistic end (0.5C/1C 3500 cycles), CALB is at the far conservative end (0.3C, 2000 cycles)
- I don't have a ton of trust in the cycle life estimates of some manufacturers. Both because I'm not sure how accurately test conditions mirror reality and because I think some push the spec's knowing it will be years before they can be verified (This is just my perspective)
- Like everything else in the datasheet, it applies to grade A cells that passed QA and are deemed fit for sale and warranty to important customers, none of us buying on the grey market know for sure why our cells are discounted, or what condition led to them failing QA. In many cases it might just be slightly lower capacity, or some other semi-benign cause. On the other hand, could their be some underlying weakness or other issue with a batch or model of cells, that leads to it being sold at a discount (quite possibly in my opinion). Therefore, even if the cycle life numbers are accurate, it seems to me that it is not a given that non-A-grade cells will definitely perform identically (particularly in cases where IR is the reason for not passing QA, or where there is some mechanical or chemical defect. I fully acknowledge this last point is highly speculative, and possibly unfounded.
Especially with that last point in mind (but still true even if you ignore the last point), I think I think your position:
Is totally reasonable and prudent if you want to get the most out of your cells. But I also think that a few test cycles down to 2.5V won't make a difference in the big picture. But based on the goal you articulated, I think you are right that a full test is not necessary for your situation.