@nagyeger: "...'Most' orbital impacts are going to be 10-14km/s..."
Low-earth orbital speed is 8 km/s, relative to the earth's center. Most satellites, and therefore I assume most dust, is in lower/medium-inclination orbits, about 30 degrees for US launches, about 50 for launches from the former USSR, etc. All that stuff will indeed be moving, relative to each other, a bit slower. (Though if you're headed across the equator going south in a 30 degree inclination orbit, and hit a speck coming north from a similar orbit, 8 * (2 * sin(30)) = 8 km/s.)
Anyway. You raise an interesting point: satellites in higher (polar and near-polar) inclination orbits, such as some recon and communication satellites, ought to hit dust particles, on average, a little faster. Therefore, they ought to see these unexplained electrical failures more often.
On the other hand... since not as much stuff is in those orbits, they'll spend a good chunk of time going over the poles, in areas where the dust is less common. That'll work in the other direction, making them _less_ likely to fail.
However, when they _do_ fail, it should be more likely to happen at lower latitudes. If an analysis of failures showed that tendency, it would support this theory that dust impacts matter. I was going to suggest checking the frequency of failures in still higher orbits, but those are above the Van Allen belts, are in sunlight almost all the time instead of a bit over half the time, and therefore have other factors that wouldn't be easy to consider.