Ermm.. maybe you might mean something else but it would be impossible to 'see the flash travel from left to right' in a single photo. Even in a series of frame it takes highly specialised equipment to 'visualise' light travelling such as this:
http://newsoffice.mit.edu/2011/trillion-fps-camera-1213
So what you're seeing is something else and without seeing examples of pics I won't speculate what it is.
Now back to the ghosting effect. I agree with you silvermoon1407 that the orientation of someone else's flash should have no effect. Of course if you're lucky enough that at exactly the same time during your exposure, someone else fires their flash then you could capture their flash pulse.
Flash pulses are incredibly short, typically less than 1/1000th of a second. But suppose you were really lucky, that pulse which may have also created the flare may have occurred in the middle of when your shutter curtain was opening or closing.
Now this would only make sense if your shutter curtain moves horizontally as oppose to vertically in landscape orientation. But I have no idea what the Nex-5R shutter curtain is like.
As Octarine had asked, do you mean the shutter is horizontal, moving vertically (when in landscape orientation)?
I couldn't find confirmation of this online. But a simple way to test is to shoot with a strobe at above the max flash sync speed without any high speed sync enable and seeing what direction the black bands are.
This is just a stab in the dark btw. I'm not particular confident in my theory but I can't think of what else that would produce this sort of 'cut off' effect.
The only other thing that even remotely resembles this is mechanical vigneting, not caused by accessories like an inappropriate hood, but typically when using a very fast lens on a smaller format camera. The image that emerges from the exit pupil of the lens is larger than the chamber inside the camera and gets clipped. But I've only seen this happen in bokeh balls and not in flaring/ghosting.
