Id say it must be lifepo4.
Re the balance after the ride, with lifepo4 in particular, there really is not much information to be gained by voltage at mid discharge, since the curve is so flat.
If you have a bms, charge back up, wait half an hour for the bms to do any discharging it might do, then try again to charge. To get the charger to restart, you might have to ride around the block, lowering all the cells enough to get the charger to go again.
If you don't have a bms, charge the lower cells one by one to 3.65v max, or at least, to slightly more than 3.5v. 3.5v is full, but lifepo4 tolerates a slight overcharge very well, so you can set a charger to charge the pack to 3.65v average, and force a balance each charge.
If inbalance persists, despite a very low rate of discharge and moderate discharge depth, then you simply have some funky cells. A lifepo4 cell that can't retain 3.3v overnight is pretty worn out. Still usable maybe, but not fresh no more.
Re your last reading fully charged, anything over 3.5v is fine. Those are balanced, even if one is 3.5 and another 3.65. This is because 3.5v is full, and any more is just surface charge that often dissipates overnight. Its the ones below 3.5v that you need to get more charge into. I call anything within .05v perfectly balanced myself. so any cell at 3.45v is fine, and typical for any pack. Even .1v off can be tolerated, if you don't need 100% of the potential capacity to do the ride you want to do. Over time, all will take less charge, but if all still charge to 3.4v or better by next morning, its still a good pack.