Hi BIG. Your post has been hanging here a bit so I'll give you some info. Keep in mind though that I do not have either of your setups.
First, I would determine whether or not your cells are protected or not. This may, or may not give you some piece of mind. Just the same, I always recommend not relying on protection circuits, anyway. It's better for the cell if the discharge is stopped well short of when the protection circuit trips, if, it even trips. At low current levels it is possible to over discharge even protected cells, for example.
As for stopping discharge as soon as there is a noticeable drop in performance, this is a good idea. Better than that though, what I would do is check the rested voltage of cells occasionally during discharge in each of your applications. This will give you a better idea of where your cells stand as far as how discharged they are. Once you are aware of the results, you won't need to check so often, or maybe at all.
Keep in mind that the voltage you read when removing a cell from a light, is not the same as what the cell's voltage was when under load. It will always read higher. The 2.50 Volt minimum voltage for the Panasonic NNP cells is when under load, not after you remove the cell from the light.
To allow for this discrepancy, a good rule of thumb (although not "carved in stone") is that cells discharged at a high rate of discharge should read around 3.5-3.6 Volts when fully discharged and cells discharged at low rates of charge should read no less than about 3.00 Volts, when pulled from the light (and also should be charged as soon as possible, not left sitting around for days, or weeks). Again, this is the rested voltage of the cell after it has recovered voltage a bit after removing it from the light, not the minimum voltage when it was in use.
Obviously, cells that have been discharged at higher rates do not become as discharged as those that were discharged at lower rates, but the voltage when under load, was about the same. That's just how it works. Voltage drop under high current loads is more than at low current loads. This is also, as I mentioned earlier, where protection circuits can fail to stop the discharge soon enough. Unlike laptop protection circuitry, the protection circuits added to Li-Ion cells are not "tuned" to any specific application, they are a "catch all" type, so can fail to do the job, particularly in low current demand applications.
So anyway, stopping discharge when performance drops off is a good idea, but may not always work when discharging at low current rates. Do some experimenting, checking cell voltage at certain points during discharge, and you should get a better idea of where your cells stand in each application.
Also, the exact figures I mentioned are not necessarily exact, just a general guideline that I use, YMMV. And, as for specific lights/dropins and such, a lot depends on the specific LED and driver etc, so gauging cell discharge level by performance alone, may vary quite a bit.
Hope this helps.
Dave
