I took a shot at a cutoff modification to see how effective it was. The light I used was a Lumapower MRV SK with the OP reflector. I used aluminum foil between the reflector and lens, like this:



The beam looked like this:



I stopped the bike on a dark 2-lane highway and walked out ahead of the bike so I could see how effective the cutoff was. It was not impressive. I went back to the bike, flipped the light over so the cutoff was now an anti-cutoff, and walked out to observe it again. There was no perceptible difference to me as a pedestrian. So much for that plan. I see people advocating cutoff-equipped lights quite often, but I don't think many have done any first-hand research.

For me, I would need to ride pretty slowly, with no competing light sources, to manage on 100 lumens (by which I mean, see and avoid road hazards, not just stay on the road). The darker it is, the less light I need, since my night vision can adapt to its fullest in darkness, but 100 lumens is a little skimpy.

The photos above, that show my bike next to a car on a highway, are from this video, for those who might be interested in the source: http://www.youtube.com/watch?v=ThwjZ6j11mc It's an out-&-back, so skip to about 1:40 if you just want to see headlights. The bike has a DiNotte 600L and a Dereelight DBS V2 with the OP reflector (DiNotte: 600 lumens and floody, DBS V2: 250 lumens and focused). I also have a 140-lumen Fenix L2D on the helmet. At the point shown in the photo, the DiNotte is responsible for most of the light reaching the camera... it's actually not much good as a highway-riding headlight, but the "shotgun" beam makes it good for safety purposes in the city because it appears intense from a wide angle of approach.