But when you turn off the buck, won't the voltage go back high? Won't it oscillate?
It probably would yes. I could slow that down with caps, or make it "AND" with the MCU. The MCU says ON (knowing the voltage) AND the zenner diode is conducting. Making the zenner just a safety cut out.
The buck turning on and off, one and off, isn't really big issue, it's basically PWM. However, when testing a buck/boost converter at 4V it was perfectly stable. At 3.8V it went nuts and slammed the output voltage to -32V when it was actually set to +12. It's that kind of response I have to avoid, hence hardware lock out.