Re: What's sixze got to do with it?
actually, a larger sized vessel would have a better positive bouyancy on the surface. You can run the calculations, but basically the displacement of the hull, combined with the volume of air inside the boat plus empty ballast tanks, and the total mass and shape of the boat itself (with empty tanks), determines how much of the boat must be submerged at all times. The larger the boat, the more likely this is to be a higher value for positive bouyancy. Typically that might be something like 85% of the hull submerged at all times, and the other 15% is controlled/compensated by ballast and trim tanks.
in this case...
It's a fair bet that there was a major seawater leak, and too much water went into the people tank, forcing the boat to sink. Blowing ballast and pumping trim wouldn't have been enough to get positive bouyancy.
I suspect that some low-price low-quality component failed (the U.S. Navy has a program called 'sub safe' to prevent the $5 part from sinking the billion dollar boat). It could have been a port hole, a valve, a weld, a pipe fitting, a ballast tank vent, or anything that couldn't be properly isolated, nor have the drain pump [assuming it had one] pump the water overboard fast enough.
All of those kinds of safety systems would have to be designed with surviving a flooding casualty.
So one possibility is that one of the portholes blew out because it was made of substandard materials. Again, $50 part sinking the million dollar boat. Oops.
They should have had some means of RAPIDLY isolating the cause of the leak [assuming there was not]. They should have a means to get the water off the boat assuming "worst case leak". The 'drain pump' would have to be reliable enough to run with the power out (let's say on separate battery power), and run submerged if needed. A backup pump would also be a good idea. Being able to isolate compartments would be even better, especially if you can pressurize a compartment that is being flooded. [watch any old sub movie, and they'll talk about that, pressurizing a compartment to keep the water out, and run on the surface so the water pressure is lower, use the drain pump to get water out, etc.]
So yeah making a _SAFE_ boat that can go underwater to 500 meters is NOT something to be taken lightly. You have to consider the behavior of materials under cyclic compressive stress, the ability to recover from a reasonable flooding casualty, safety systems that can be remotely activated, 'emergency blow' on the ballast tanks guaranteed to work every time, and operating procedures that go along with all of this stuff [as well as maintenance]. And if one of the ballast vents fails, can you recover from that? How about 2 vents? Compartmented ballast tanks with multiple vents helps make that possible. I think the old WW2 boats had 6 or 8 ballast tanks, each with its own vent valve, that was shut as soon as they submerged so they could emergency blow on a moment's notice.
And all of the welds on the hull should be x-rayed for cracks and other defects on a regular basis. The fracture toughness of the hull material should be well known, and all design margins calculated based on the minimum detectable flaw sizes, like would be done for a bridge, or a cargo ship, or anything else made of steel that's likely to undergo heavy stresses during normal operation.
And the maximum allowed depth should be based on the worst case design margin in the worst possible place, for safety. The claims of 400 to 500 meters is pretty optimistic, yeah.
/me points out that at 500 meters, water pressure would be about 800psi... imaging getting hit with a stream of water at 800psi because something broke. Even a half-inch hole at 800psi could slice you in half. It's no joke dealing with this kind of thing.