I remember seeing that experiment, but it did not address the scenario that I have presented. I am not talking about flooding all 16 compartments evenly; I am suggesting keeping the watertight doors between compartments 4 and 5 closed, and keeping all compartments aft of 5 open, or at least, slightly opened so that the rate of flooding of compartment #5 would be considerably slower, thereby lengthening the amount of time it would take for the bow to sink low enough for the water flowing in from the damaged forward compartments to cross over the bulkhead between compartments 4 and 5 to increase the flooding rate aft.
It would take some very complex differential equations involving rates of flooding, rates of flow through doors only slightly opened, and rates of pumping out the water to determine whether this could have saved the ship, or at least bought more time. But it should, I think, have been a consideration since it would have slowed the flooding of the 5th compartment, and possibly bought enough time to staunch the flow from damage in the 5th compartment if the water level were not able to increase too quickly above the level of the door sills as water had to flood up to that level in each additional compartment moving aft.
If the pumps in boiler room 6 were keeping the water at bay for its sustained damage, and there was no further damage aft, then opening the doors to the undamaged compartments even slightly from the flooding 5th compartment would lessen the rate of sinking and possibly reach a point of stasis between pumping and flooding if the more powerful aft pumps were able to be brought into use before the water flow from the damaged first four compartments increased the rate of flooding of compartment 5.
Of course, once the bulkhead between compartments 4 and 5 had been breached, the full flow rate of flooding would ultimately sink the ship.
Perhaps, by the time Andrews had assessed the damage, the bulkhead between compartments 4 and 5 had already been breached. If the ship made 14 feet of water in 10 minutes, then it would only take 25 minutes after the collision for the water to flood to a depth of 36 feet in the damaged compartments. Perhaps my theory could have worked, but ultimately it would have had to have been implemented almost immediately after the collision - an unlikely idea before the damage had been completely assessed.