check out: http://www.youtube.com/watch?v=MN32cH9LnzE
There are huge discrepancies in the speed estimations, in the description of the impact angle and flight path. At the same time there is very little description about the way these data were analyzed.
The high decorated MIT described it but used a method that added failure on failure (horizontal path, straight path, POV rectangular, Doppler effect…). Every conclusion was the assumption for the next conclusion (and so on) and increased the error of the result.
Finally, the MIT result of 503 mph had an error of about 100mph. The NIST result was about 50mph better but still bad. Btw, they estimated the overground speed of a straight line – no wind, no descend, no turn… The reality looked a bit different.