Several people have commented on the final points in the FlightAware track log and their reliability. The last three points are interesting - the first two of these plot horizontally exactly on the extended centerline (as they should), but the final one plots about 3,000 feet beyond the aircraft's final resting place (on the centerline of 1R/19L).
The second to last point is about 750 feet from the sea wall, and the ROD is 120. The time interval between these positions is about 10 seconds. The final point would have been recorded after the tail strike on the sea wall, which would have occurred about 4 seconds after the next to last point was recorded. Therefore, the 120 fpm rate of climb given for the final point is meaningless, as is its altitude - the aircraft was already on the ground. The reason for the erroneous lat/long for the final point is unknown. The FDR data will resolve all of this.
See these posts and compare them to the Google Earth image of the runway:
http://www.pprune.org/rumours-news/5...ml#post7926888
http://www.pprune.org/rumours-news/5...ml#post7927198