If I may add a comment or two. The L2 low level is certainly much brighter than an output that you will get from the KL4 with a resistor switch set up. I frankly think of the L2 as having a normal level and then "over drive" from the high. Regardless of one's perception on these levels, you won't get the L2 lower level from the KL4 knocked out of regulation by the resistor switch.
In terms of useful low levels, efficiency and run times, the comparisons are not so straight forward as one might think. I leave it to the electronic guys to sort out but even though the L2 is regulated on its first low setting, you do have a 10 ohm resistor that is dropping the voltage to get you into low. The power wasted by the 10 nohm resistor is part of the loss in efficiency on the L2. On a KL4 with a 22 ohm resistor, you are wasting the power in this resistor as well but how does the efficiency compare? /ubbthreads/images/graemlins/icon3.gif
I posted a
run time graph that is misleading as you see a sharp drop in the curve. This drop is due to the ambient lighting changing. I turned the shoplights off and went to bed! /ubbthreads/images/graemlins/tongue.gif In actuality, the first 4 hours of the graph should be dropped down to join the rest of the graph. You can see that I got about 11 hours of runtime until 50%.
The resistored KL4 is a very useful and viable solution but it is not the same in terms of light output as the L2.