Why do you think it would work any differently? So DumpsterDiver is diving with a new or not his usual buddy. That buddy signals OOA. You give him/her your primary and start breathing off your AIR II and up you go. Now let's say DumpsterDiver has an octo instead and it's bungeed around his neck. Same buddy signals OOA. He donates his primary and starts breathing off his octo. Up you go.
How does the fact that his inflator having a mouthpiece on it just screw it all up and confuse the OOA diver? If the OOA diver has never dove with someone who has an octo around his neck, will he go nuts and get confused? Will it cause even more panic because he has never seen this set-up?
The point is that posting a video of someone calmly doing a skill for a camera on an ascent line doesn't prove much about the gear or technique, that was my point. Maybe DumpsterDiver wouldn't have a problem in a real emergency - he has a ton of dives. But maybe someone else using the Air2 with only 20 dives might, especially when the emergency is unanticipated and real. That's what this discussion is about, that's all. If you really think that there is no difference between the execution of procedures in a drill and in an actual emergency, then I would have to disagree with you, and leave it at that.
Look, gear optimization is always about tinkering around the edges of design, trying to look for minor modes of failure, etc.. It is almost NEVER about having an accident analysis report that shows how this or that device killed someone. With rare exceptions, we don't usually find out what happened in a fatality or serious DCS incident. But that doesn't mean that it's not worthwhile to discuss things here.
None of us has enough experience to actually do a real scientific analysis out of our own log books, so it's always going to be conjecture. And conjecture isn't a bad thing. Hey, we're not diving today anyway, might as well talk about it!