Why are you going through that restriction if you're on your backup already?
Ok, ok, you got through it on the way in, THEN had to go on backup. I guess I can see this, but this looks to be one of those "I have to accept that double redundancy is simply not reasonable" things. And you are talking about a VERY nasty restriction if you can't manage to get to one of your posts. Diveable, yes, but man, that's tight.
Push come to shove, assuming you're using stage/deco bottles, and leave the second stages finger tight, you could swap one before going through the restriction and restore your primary. Is that risk worth it? I don't know. But I'd definitely think about it if I was in that situation and knew that for the length of the restriction I can't reach back there to either insure the valve has not rolled off, or to turn it back on if it does! For that matter you could go on a stage bottle for the length of that restriction; you'd probably have to dismount it anyway and push it ahead if it really was tight enough that you couldn't get to your left post. Of course that option doesn't apply for deco gas (especially if you left it, as you might in a cave. You probably know why that's a bad idea in a wreck, right?
)
Note that many of those older manifolds, while they could not roll off the left post, were also all but inaccessible in terms of the valve access for the left post as well. If you can't get to it at all, then that's no good either, as the far more likely risk is a first stage failure that dumps your gas and requires immediate attention.
Personally, I think the better question is what kind of failure rate, if any, does a barrel O-ringed crossbar have, and what are the odds of a catastrophic (both sides) failure that dumps BOTH sides of the isolator.
You have to look at these things as a potential failure rate for the entire system. Manifolded doubles are an attempt, primarily, to protect against a first stage failure. To do that effectively the valves for both posts must be EASILY accessible, lest your dump the contents of both tanks before you can close the post.
The isolator is an attempt to protect against a burst disk or tank neck O-ring failure. However, it adds two O-rings (per side) that can fail while protecting against two things (neck O-ring and burst disk) that can fail. Is that a good trade-off? On balance, and all on its own, probably not.
But the isolator gives you protection against one more thing - a catastrophic post failure that dumps the gas (e.g. a post that strips open and leaks, threatening to dump the entire contents) in that you can protect half the supply this way, albiet at a loss of all redundancy (you lose one first and second this way too, and one source of inflation as well at the same time)
I've thought some about this, and I believe the isolator system, while introducing some risks, alleviates more than it creates. The restriction issue can be managed provided this is a staged (or deco gas carried) dive, as you do have at least one extra second stage with you.
