Depends what you mean by “struggle”. Usually, once you go above 3.45V per cell, there is very very little additional energy storage to be had. Above 3.6V, even a small amount of current can cause a very rapid voltage rise, and cells won’t stay at this voltage for very long once you stop charging them. So depending on what the struggle is, this might not be a reason for concern.
What is a reason for concern, is that obviously this cell behaves different to the others… right?
Much better! 
I recall you stated that cells from the same batch etc. shouldn’t go out of balance once they have been balanced…
Am I correct??
I think it is a little more nuanced than that. In all likelihood the manufacturer uses a “binning” practice. The cell capacity is tested and those of similar capacity are thrown in the same bin. Those with a lesser capacity gets thrown in the next bin, and so on. And if you assemble a battery from cells that come from the same bin, they tend to stay in balance with less effort.
But because (on a microscopic scale) no two cells are perfectly the same, nor do they age the same, there will always be small imbalances. But because you started with cells that are very similar, you don’t need a big humongous balancer to keep them in balance. 50mA or so will be enough… 
Depends on how the bank is used, and if it frequently sees 100% SOC (3.45v per cell for DIY) where the BMS can balance the cells, as balancing starts on wot, 3.375v or 3.4v depending on settings, it will go slightly out of balance, perfectly normal, the BMS will sort that … with some patience, balancing at milliAmps.
The secret lies in Batched Matched (same capacity and resistance) Grade A (or better quality) should ideally be used on a daily cycled system. Most brand names use top-quality cells.
But, if cell/s go badly out of balance, a cell/s has a problem, as SolarIsDeWay has found. Maybe it can be fixed.
Few tricks one can pull under DVCC, like dropping the charge amps first, to help BMS balance faster on SOC above 90%. Even drop the volts, to help the BMS mA balancing to catch up.
Tibit: 30/50/160/270mA balancing on say like 16 x 280ah cell bank being charged at 10’s of amps … ok.
So one needs to get it right at the start, the cells having close as damn the same capacity and resistance over all the cells, or you will have to pull a few tricks and if you are lucky, even find the “sweet spot” of the mismatched cells. But it is a temporary solution.
This is how it looks, Batched Matched Grade A cells:
Just to clarify these voltages are for LiFePO4 cells…
And yes I agree that Li-Ion aren’t forgiving like lead-acid batteries.
So not only do Li-Ion batteries require a BMS for over/under voltage but they also need a balancing cct.
I’ve come across 18650 battery packs without the latter which causes the battery pack to perform less than full capacity. Has anybody else come across this as well??
Necro-ing an old topic, but some new info came about after I dug into a US2000B earlier this week. So on this bit…
What I learned is that on that balance connector, you will find a white wire in the right-most position. That is the negative side of the bottom cell. The 5 wires that follow after that (to the left) are each attached to the next junction between cells, and should show deltas of around 3.4V-3.5V between each.
If you attach dupont wires to each, you can carefully balance the cells without taking the whole pack apart.
The remaining wires on the connector are – this is an educated guess – most likely for temperature sensors.
thanks for this…
I still struggle with cell 10…charges slower than the rest and discharges faster…causing HV/OV for other cells at about 65% SOC.