What they did in this paper is measure the membrane potential along a mitochondrion, so along its longitudinal axis. They did this for about 80 mitochondria.
What they find is that within a mitochondrion, the membrane potential is more or less the same, the variations are around 5 mV. Interestingly, the points at the ends of the mitochondria were usually (in 78% of the cases) extrema. So it often occurred that the potential at the ends was either the minimum potential along the whole mitochondria, or the maximum. This maybe could have something to do with the ends being involved in fusion/fission processes.
They also find that the fluctuations in potential in mitochondria that are connected in the network occur simultaneously.
Another interesting thing is that they find that the potential does not depend on where in the cytosol the mitochondria are. This sort of contradicts other findings that the membrane potential depends on the localization of the mitochondrion in the cell (see e.g. http://www.pnas.org/content/88/9/3671.abstract)
Another thing that they mention (they don't measure this themselves) is that even if mitochondria don't have DNA, they still keep their membrane potential intact by using their ATP synthase in reverse. The ATP that is needed for this comes from glycolysis. But now the mitochondria are just things in the cell using up ATP and not generating anything at all, so I was wondering why they maintain their membrane potential at all. I guess the cell doesn't want to kill all its mitochondria if they don't work for a while..