Well, whether you notice it or not, there is a difference. CR123As are not rechargeable and measure 3 VDC as apposed to a rechargeable Li-ion battery that will measure 4.2 VDC when fully charged. That is a full 1.2 volt difference, and the 808nm drivers won't output as much power using CR123As.
Most of the green drivers I've seen are LDO. With a Vf of ~1.9V, what makes you think 1V or so isn't enough headroom for a LDO? Some of the green drivers I've tested go all the way down to 2.2V while maintaining their current regulation.
I have a couple of these lying around somewhere, so I'll have to test that. Unless the drivers in these are somehow worse than the other chinese stuff, I suspect you're quite wrong.
My experience with these cheap 532nm lasers like the 301s and 303s is the drivers in these are not as good at regulating current as well as one might want. For that reason, I have found the output is determined by the input voltage to these cheap drivers. It has nothing to do with the Vf of the pump diode.........just a noticeable difference in the power output when driven with these ubiquitous drivers. Also, the OP is using it for a 405nm and the boost driver in that circuit might be close to cut off at 3 volts.
nice data cyparagon. so if some drivers are affected by the lower voltage and some arent... has anyone tested something similar with survival lasers 1.25A driver? thats the one that will be in my 638 build and was wondering if the difference in voltage would affect that driver much. im gonna buy myself some 16340s on amazon either way since my local battery store only sells rcr123s and the vape shop nearby overinflates the prices on their batteries.
Good drivers should have no effect as they regulate the current to the laser diode. As long as there is enough voltage to drive the diode, it should remain constant. If the battery doesn't have the current to drive a diode, then I'm not sure if the battery voltage will sag, or the driver just won't pull the current and stop regulating.