The study you mention about green light is measuring light in a way I would not recommend - all the measurements are normalized to "luminance". Because they measure in cd/m2 and not equal photons, I believe they are testing 20x as much blue light as green, so it is not conclusive at all - if you are using 20 times as many photons you'd certainly expect it to make pain worse.
We have more work to do in understanding when people experience the greatest light sensitivity due to migraine, which is why there is no schedule for headaches in f.lux right now, even for removing blue light.
The few studies that have looked at the times of migraine suggest that the first half of the day is the most sensitive, but other people report that it ramps up later in the day.
If we can understand how to automate it, we will, but for now perhaps the right answer is to make the manual controls easier to use in the meantime.