and the consequences for those that need to quickly turn the valve off?
I'm not sure why you assume that a knob which has a rotation lock would take more time to close. That would really depend upon the design, but in my opinion a requirement of the design would be that it would not noticeably increase the time taken to manually close it (hence why I do not like the zip tie suggestion), would still be easy to operate using a single hand wearing thick gloves, and would not reduce the physical strength or robustness. I don't think those goals are unrealistic...
Is this problem *of which I am not aware of any documented deaths, only near misses*, worth risking the lives of those that need to turn off quickly for feathering *of which I can think of several where the several second delay in being able to shut the valve off could very well have killed the diver*.
If someone wanted to feather it, then they simply wouldn't lock it open. Anyone who's advanced enough to feather the knob can surely handle a single additional motion to unlock it first.
Another issue with this type of valve is that when doing things like ice diving, there is a high propensity for valve/regulator components to freeze. This type of valve if frozen while open would be completely unable to be shut off
We haven't even discussed the mechanism of the lock, so you can't really claim that it would be completely unable to shut off when frozen..