Light travels independently of the train, at c. The time it takes the light to travel to the train observer's midpoint position is different for each strike. It is true that the train observer remains at the midpoint of the train, but when opposite ends of the train have light sources, and the lights are simultaneously activated (easily done with two lights and a common switch at the midpoint, activated by the observer), during the time it takes for light to travel the half length of the train, the observer also travels away from one light, and towards the other. That creates a situation that the light from one source has less distance to travel than the light from the other source, until it impacts the observer. I've already showed why synchronized clocks will appear different to the observer on the train, which is a red flag as to the observer's actual velocity, unlike the embankment observer's actual zero velocity. You can not say the train is at rest and that it is actually the embankment that has a velocity (or any other combination of the two), due to the fact that the strikes occurred as the points lined up, and the light from the strikes impacted the embankment observer at the same time, and the train observer at different times. That is not a reversible situation. The points MUST be lined up. I explained the difference with the clock phenomena, which is also a true indicator of a zero velocity.