You can't test longevity without testing the diode's limits.
As stated earlier, the 405 and 650 diodes don't handle a beating well. Once they LED, they're gone for good. This test will likely end either when that happens, or when the power increase starts to plateau.
We will never know longevity of a diode until it's been used for a while, or you get a beefy heatsink and cooling method to run the laser constantly for a long period. Either would work, but the first step is to test the limits of the diode. Without that information, we can't go anywhere.
Welcome to new diode testing
Baby steps. If it proves to be a decent diode, more will be bought. Then perhaps we could do a 16X graph like we did with 8X and 12X diodes. Where the builder adds all the stats of that build, and when it dies, they report it. That will tell us longevity.
One step at a time
EDIT: Even after all that, some diodes just do whatever the hell they want *cough* GGW 6X *cough*. Those were known to work wonders at higher currents, or insta-death at currents below what was considered "safe". Some diodes just are not made to handle above what they are rated for, at least not consistently. I knew of people going through 2 or 3 GGW's to get a build at the current/power they wanted. To put it bluntly, shit happens.