I'll trust what Jason says, but I'm not convinced that you've understood him correctly. Can you point to where he's said this for me? The reason I question it is that if it's true that it continuously balances regardless of state of charge, then what is the purpose of "triggering" at a high state of charge since it is always possible to measure the cell voltages at any time?
In other charge balancing systems that I've worked with (all much smaller packs with no more than 15 series cells, but the principal is the same), the purpose of balancing at a high state of charge is because the charging currents are necessarily small then and you can use practical, small, loads for the balancing while charging and have the pack balanced when you've finished charging. But they only do this balancing while charging close to the upper limit of capacity, just as the Tesla is reported to be doing.
If you "bleed energy" over days then you are wasting stored energy in the cells which is inefficient compared to having not put the energy in the cell in the first place. You certainly cannot allow any cell to be overcharged, and then bled down later to match its lesser-charged peers, so you are bleeding energy from cells that aren't even fully charged, rather than allowing under-charged cells to get more fully charged. That makes little sense. Instead it is normal to simply bypass the charging of the more fully charged cells (with a load that doesn't discharge the cell, but instead provides a path for the charging current to the other series elements to flow through the load instead of the cell) while allowing the lesser charged cells to catch up.