A few things noticed: MCAS seems to be a fast_put_together SW solution to a certification problem (stick force/g). Not a stall preventer directly. It also looks like B forgot to include data validity checks and "reasonables" checks that would return any trimming actions done after AoA returns to safe range.
A thought I had when reading what STS does, that sounds like SW augmented longitudinal stability. Meybe B thought that when that was acceptable other similar patching would be ok too.
Maybe they really should retire the 737 base and have a new one (yes, that does cost a lot), design methods have advanced a lot since the first flight of 737, which I understand was a good airliner in its first versions.