The AoA sensor as such may deserve a closer look. First, it is difficult to google their physics and history in the first place. There is a NASA report from 2014 summarizing the historical development. Seems that these type of sensors are relatively new outside military planes. Feeding its output into automated systems may not be straight forward: “Sensor Fault” may come in many shades. As a good sensor, it certainly has DC voltage lines monitoring and AD well-being indicators and possibly an automated test signal insert with check functions. So the case of using several sensors, like in a Sub-Prime bond bundle, versus the good caring of one deserves attention.
With all this considered, it seems difficult to judge a system that takes a non-monitored output from one complicated sensor to drive a control loop.