They don't even MENTION the filter on the DFDR in this program, even though the NTSB Hearing in October, 2002 began with the words of Bob Benzon himself: "Due to the filter that augmented the readout data from the DFDR, we will never know exactly what the flight controls on AA 587 did on november 12, 2001" or words to that effect. He then explained that, the DFDR normally records 60 data points per second. The filter installed--which was ordered by the FAA to be de-activated years prior, but was not--filtered out 59 of these datapoints, so that only one data reading per second was sent to the cockpit instruments.
In order, however, to get a good read on the rudder, the filtered-out data points had to be put back in. Two models were made of this data to re-introduce this...and, the project team called in to perform BOTH models?
Airbus.