As Mr Happy has pointed out, besides charge termination, all of the other areas are kind of grey and are application specific and depend on the number of cycles desired from the cell, the load on the cell, variations in manufacturing methods from one brand to the next (some are more resilient (spelling?) than others) etc etc... There is also an issue of time, an over-discharged cell that is recovered very quickly after the fact will not suffer as much wear-n-tear as one that is stored in a discharged state for a long period of time.
Here's some rough examples:
Lets assume a cell under a light load, like 0.2C, in a well designed laptop. (5 hour runtime).
The cell will charge to 4.20V, then under a load will start off pretty close to that, we'll say 4.12V, as the cell drains it the loaded voltage will drop to somewhere in the ballpark of 3.5-3.6V (give or take) and most of it's available stored capacity at this point will have been exhausted, the cell will soon take a nose dive in voltage and the laptop will probably shut it down pretty quick like at this point, he device might allow the cells to go to an average of ~3.2-3.3V before it would more than likely shut down because a well designed device knows that the cells are already in the process of face planting. If the cells Were tested after being terminated at ~3.3V, they might read 3.5-3.6V open circuit but would be considered basically empty.
Now lets crank it up to a 2C load, (a pocket rocket flashlight with a 30 minute runtime).
In this case, obviously the cell is still charged to 4.20V, but as soon as that intense load hits the cell, it can immediately drop to ~3.7-3.8V. The cell will be *basically* dead at at somewhere around 3.0-3.3V depending on the behavior of that particular cell, and the low-volt cuttoff on such a device might be something like 2.5-3.0V, after the load is removed, it should rebound to 3.3V or higher ideally.
---------------------------------------------------
If the cell does not rebound above 3.0V after being removed from the load, then it is a safe bet to call it over-discharged to some extent. Sometimes a cell can in fact be pulled down to 2-2.5V or so under a heavy load and then bounce back to 3V after about 30 seconds of resting with minimal damage. I would recommend against it, if the load is harsh enough that the cell is operating down in this range under a load then it's probably not great for the cell to begin with.
I like to see 3.5-3.7V or higher open circuit after a discharge regardless of load, this is a great place to call it quits and will improve cycle life.
---------------------------------------------------
Technically speaking, the best state to store a li-ion cell is partially charged. like around 3.9V is a good place for it. The longest shelf life and the least internal break-down occurs around here, The farther away from this point the cell is stored at, the worse off it will be. Personally, I like my cells charged in storage because I want them ready to go if the power goes out, so when they wear out I just have to replace em and deal with it. Storing below 3.5V for extended periods of time would definitely be hard on them, letting them sit at 3.0V for a few months would probably nearly destroy em, anything less than 3.0V would pretty well do em in in a few hours/days/weeks depending on how far below 3.0V they are.
