I even threw together a primitive optical bench and measured one, but I can no longer find the results. It’s posted here (or in the archive) somewhere, I think.
But bear in mind, the pulse output is not a pulse output - it is an over-driven analogue output, so you can get a linear response over a limited range of light levels.
I cannot remember now where the LED on the back started to light in relation to the output voltage - hence in relation to the voltage at which the pulse input recognised a pulse.
Here are the results: First try with EmonPi - Pulsecount stuck at 1 - #16 by Robert.Wall
(This is for one sensor, yours won’t necessarily be the same.)