In theory, if we change reflector and leave everything else constant, beam intensity is directly proportional to reflector area. In practice some reflectors are better than others, but the best ones out there show similar cd/lm/mm² regardless of size.
Examples? Let's list diameter and cd/mm² of several optics with XHP35 HD driven to 2700 lm.
They are all the Pareto-frontiers in term of throw vs. size.
LedNLight LLC02N, 16 mm, 48.3
Khatod PL19706, 19.7 mm, 55.6
Carclo 10199, 20 mm, 54.1
Carclo 10048, 26.5 mm, 52.4
Khatod PL119806, 26.8 mm, 53.2
Carclo 10755, 30 mm, 52.7
LedNLight LLC05N, 32 mm, 53.4
Khatod PL60006, 35 mm, 55.9
LedNLight LLC49R, 45 mm, 54.7
LedNLight LLC56N, 67 mm, 55.0
This is not very close to what we see here as these are all TIRs. And the LED is not very interesting. But I show this because that's what I have data for. The pattern is clear - performance is directly proportional to area,
That's why I care about cd/lm - given the same LEDs in several hosts it allows determination of reflector quality.
There are some problems with it though:
* head diameter is not reflector diameter. We often have to estimate the latter
* size of LED opening matters
* aspheric-based lights are different beasts and have different active areas, making direct comparisons harder.
BTW for me Manker U22 is not a dedicated thrower. I intend to use it as a general purpose walking light, Though not with Blackie.