This actually seems to be a routine occurance around the world, especially in the third world, for large quakes; the country it occurs in puts out this really low mag estimate and NEIC is much higher (and NEIC turns out to be right 100% of the time.)
Basically, what happens is that these really large quakes "saturate" or "peg" the local seismographs; if you have them set to be able to pick up and catalog your average quake a really huge one will basically cause the "pen" as it were to go off the side of the chart.
The other issue is the complexity in calculating the magnitude of large quakes. There are 3-4 different scales (Ms, Mw, etc....interestingly enough, the "Richter" scale was only designed for Southern California, and is no longer used anywhere, but the media keeps using the Richter term). You could see NEIC revise the estimate several times in the next few days. And you could see scientific papers arguing over the actual magnitude months from now.
Thanks for the information!