Smythe, et al.
There was no need to change the basic MCAS algorithm. It was ‘tried and tested’, and certificated.
The safety issue involved the integrity of the AoA input to the computation, where an error resulted in unanticipated trim operation due to weaknesses in the implementation - design good, engineering and certification not so. Without changing or improving the AoA vane, the initial solution (1) concentrated on limiting the effect of an erroneous value; cross comparison of sensors, limiting the extent and rate of trim.
Apparently this did not meet with regulatory approval, thus a revised version is being considered.
Most system testing and validation could be completed with ground simulation / stimulation of the overall system. A test flight would validate the integrity enhancement with ‘simulated’ in-flight failures. Also it would be expected that the certification authorities require demonstrations of flight handling with MCAS disabled and any limitations in operation.
The extent of the flight tests is surprising and more likely to be related to other issues arising from the accidents; particularly the trim runway drill and the apparent inability to use the trim wheel or elevator to pitch the aircraft.
This might involve additional modification and flight testing; see tech discussions earlier in this thread.
1)
https://www.boeing.com/commercial/73...e-updates.page