Unless you can get the manufacturer to tell you exactly was metal was used and how the tank was designed (wall thickness, head design, etc) and you hire a Registered Professional Engineer, nobody on the planet (other than DOT and the manufacturer) can answer this with credibility. Well, maybe with destructive testing of enough samples it could be worked out but who wants to take the liability of saying "Hey, that 2400 psi tank is good for 4000 psi forever and always!" and then stick behind it with liability insurance.
The manufacturers are required to perform destructive testing on a particular schedule. But I notice that they don't publish the results for mere mortals to read.
Now, if you want to try to outguess the engineers that designed this stuff, you're going to be on your own.
I'm an Engineer*, and there's no way you'd ever stamp something like that and take liability. Your legal responsibility lasts for six years after your death, and frankly, something like this is just not worth the money. I'd respond with "buy double steel 130s or a rebreather." There are tanks on the market that will support any given PSI. You may not be able to lift them, but that's another issue.
You've got the right idea WRT sample sizes and destruction. There is, of course, a safety factor built into the tanks. Let's call it X. (I don't know what it is.) That would mean that your tank could probably be filled to 3000 PSI times X before it explodes.
However, that doesn't mean YOUR tank will fill to that pressure. It means that a
typical tank without defects would be able to withstand 3000 PSI times X. Modern manufacturing reduces the number of bad tanks in the batch, and testing weeds out the worn-out tanks.
The tanks are cast, which means that defects in the material or manufacturing will be unknown. Generally, you'll be okay, but you might have a little bubble near the wall, weakening the structure on your particular tank. You now have a tank in the batch which is weaker than the others by some factor Y. If your math was good, then all instances of Y will be smaller than the safety factor, X.
The reason you don't ever publish the testing results is pretty obvious. If it turns out that Fictional Tanks, Inc, publishes that their 3000 PSI tanks have a safety factor a 5, someone is going to think that means the tanks can be filled to 15k. Thoon! Then you've got someone suing Fictional Tanks, and if they're found to have 1% of the liability, they could end up having to pay for a new dive shop and/or a wrongful death suit.
Overfilling a tank is, simply, irresponsible. Yes, you'll get away with it. No, it's not the right solution. You're trusting people who aren't aware of all the facts and saying, "it's fine because I've been doing it for 20 years / everyone does it" without realizing how lucky you really are. Just because the hydro tests fill the tanks to 5k, that doesn't mean it'll work for your tanks all the time. They do wear out, although the filling / emptying process is remarkably non-stressing for the materials involved.
Yes, I know you've got a tank that you've been overfilling for 20 years. That doesn't mean that all tanks can be overfilled for 20 years. You just got a good one that's probably an outlier.
*Electrical, EIT.
