In shallow, clear water on a bright day both the HUD and controller can be hard to read, the ambient light is just too bright, or the glare can be high.I am guessing we will find out the o2 was shut off, but all the talk of checking the controllers -- wouldn't the HUD usually notify the operator with at least red indicators? The controller knows the depth.
My primary issue with shutting of O2 at depth, is what is it actually testing? If the diver is paying the minimum attention to their pO2 to notice it isn't maintaining? That is drilled over and over and over if they are driving the loop manually. It certainly isn't testing the ability to turn on the O2 at the surface, which is when the fatality is going to happen, the conditions are just not the same. If a diver jumps in the water with their O2 off they have already had several failures before even getting wet, I don't see how shutting down the valve at depth does anything for that.
I don't think that it is overly dangerous in general, it is just overly dangerous for the efficacy of the learning outcomes. Much like doing couch testing of a CO2 hit, it is dangerous and doesn't provide any useful feedback.
-Chris