I know we have looked at this before. I understood there to be consensus to the effect that mark 3 was caused by coming into contact with blood from wound 2 while the head was titled forward.
In the lower initial wound the staining can also be seen but to a lesser degree. Since her head was moved forward after the second shot and since no police officer has admitted to moving her, this would leave only one explanation.
I was just going to post about this being examined in the past although I think on the earlier forum (or even the blue), which I could have missed having come late to it.
I'm not sure that Sheila's head would have moved enough after the second shot for the wound to smear, unless it was deliberately positioned or lifted up by the assailant. According to the pathologist she would have been "stunned" after the first one. Looking at her closed eyelids I'm convinced she was asleep, had fainted or was comatose before the second shot was fired.
There's a similar photo somewhere with a line drawn on it showing the division between upper and lower smears, and they do appear to give the same effect you'd get with a Rorschach pattern using ink on paper. But since you put this up again I've had a long hard think to see if there could be other explanations. I'm not saying definitely that any policeman touched her neck to examine the wound more clearly to cause them without even thinking before any CS photo was taken, although I wouldn't rule out the examining doctor, or more likely the moderator end being played tentatively on her neck when the assailant was lining up the rifle to make sure the second bullet penetrated her brain. If the moderator's end had been coated with blood and was still moist from the first close/near contact neck shot as a result of backspatter (or whatever you want to call it) then I think it's possible this could have made the smears. Both 2 and 3 have different outlines... look at the bottom line of 3 - it's straight, which doesn't match the top line of 2 for it to be mirrored. And similarly the dark blobs at the end of 2 and 3 don't match position-wise for the smears to be mirrored.