Sure, but it also helps to know your gear and to use that knowledge to take the best course of action.
For example, a full AL80 with *no* regulator attached takes roughly 2 minutes for it to completely drain:
http://www.undercurrent.org/UCnow/articles/SpareAir9902.shtml
Since a 'burst o ring' is a more constricted airflow path its going to take longer than this to drain. At only 50% constriction, it would take twice as long - - assume that that lets us account for less than full tanks - - ie, a tank that has its failure when its only half full (1500psi) will still take roughly 2 minutes to drain.
A Rec OW ascent from 90fsw up to the surface in 2 minutes with a remaining air supply is not an ESA suicide run. In fact, in the "old days" before we had safety stops and reduced the 60ft/min ascent rates, a
completely normal ascent would have only taken 90 seconds.
So if my buddy's not immediately there, my recourse would be proceed up without delay. Sure, if I can get the valve partially closed on the way, great, but it can wait until I'm to at least safety stop depths before pausing to work on it and to try to get in a short safety stop.
True, but with such a procedure, you're adding a lot of complexity, which reduces its probability of successful execution.
It also merely swaps risks - - the on/off/blow/go bit effectively increases your risk of an Embolism for what gain? A reduced ascent rate to reduce DCS risk.
But for an Rec OW dive, which is more dangerous: a DCS hit or an Air Embolism?
Answer: the DCS hit is significantly more survivable. As such, increasing your Embolism risk just to lower an already low DCS risk sounds like the wrong risk trade-off to make.