Jake, you're using a rkcstr driver.. Two Li-Ions are hardly enough to keep it in regulation for half the battery capacity, remember? The current is only correct when the batteries have more than 7.7V.. Two full Li-Ions have 8.4V, empty they have 6V, so the current and the power start dropping as soon as the battery voltage drops from 8.4 down to around 7.7V.. This means, the batteries would have to be recharged before they even drop to their nominal voltage. For optimal battery use with a PHR and a rkcstr driver, you need three Li-Ions. If you select batteries so that they still have more than enough voltage when they are empty, the power of the laser stays the same all the time, untill the batteries are completelly flat.
But this has nothing to do with the measured power. When the batteries are completelly full, the current IS what you set it to. No more and no less.
The other laser at 140mA could really have been 140mW.. I have seen one diode so far, that could do that.
But 140mW at 100mA is weird..
Jake, could you perhaps replace the tailcap with your DMM in a high current measuring range (over 2A, 20A preferably), and report what you measure? EDIT: With FULL batteries please! And make sure you also get 140mW while doing this..
With a linear driver, the current from the battery is the same as the current through the diode. This would answer the question about your misterious diode.