Now, I will discuss the top of the Pipelauncher (the "capped end"). At top of acceleration stroke there could be 60 psi throughout the entire length of the pipelauncher (unless sections of central chamber can be sealed off- though each of the 3' would probably need to be segmented to regulate pressure differences). Now, let's look at a pipelauncher at rest. It floats by displacing a certain amount water, if we look at the 800' pipelauncher with it's weight of 750 tons, it forcing 750 tons of water below the water level which is equal just over 40' and is over 15 psi of pressure inside the central chamber. If you make the pipelauncher have a greater weight- say load it up with 750 tons of sand or water, then in order to float the water must be forced further down the chamber in the pipelauncher and be at a higher pressure: +80' and and 2 atm. Now, what happens if you drop this added weight?
As you recall the pipelauncher needs around 1 atm to float, so the pipelauncher will go up to 160' and because of inertia will go above 160' level then stop, fall back below the 160' level, then rise again- etc. Now the pipelauncher happens to have an easy way of adding weight- the 12 3' diameter pipe can be partially filled with water, and using gravity and/or air pressure this water can be expelled. As I mentioned before 6' of length of this 3' pipe has the volume of about 42 cu feet and in a 100' section you have 1200' of 3' pipe, 42 times 200 is 8400 cu ft of water which times 62lbs per cubic' is 260.4 tons. So you need roughly 300' section to give you this 750 tons- though this 300' of pipe would be needed above the waterline to give weight. Now, in my calculation for the pipelauncher's bouyancy I counted this pipes as being empty- AND at the top of the acceleration stroke they will be empty- but before this chunk of steel jumps out of the water these pipes will actually be partial filled with seawater and this would actually "subtract" form the stated bouyancy, thereby requiring more bouyancy from central chamber.
Now let's skip ahead a bit. The rocket (let's say this payload weights 500 tons) is on top of the pipelauncher. The rocket has a waterproof jacket which is 50' high. The top of the pipelauncher (where it mates with the rocket) is 150-200' above the waterline. The rocket is fueled and is a go for launch. 20 second countdown begins. The pipelauncher drops and starts to go back up when about 1/2 of the rocket jacket is below the waterline. The pipelauncher would drop by reducing the volume of air inside the pipelauncher so that it floats at rest at about 50-75' above the waterline (it goes pass this point because of inertia). Now without any added power the pipelauncher would rise back up to around 100' above the waterline, though you could gun it for faster acceleration.
Now, point of skipping ahead is to point out that for your acceleration phase you want the pipelauncher to float about +50' above the waterline. And since without any payload and all the 3' pipes clear of water, the waterline inside the central chamber would be +40' below the ocean waterline. So with a rocket payload and with the 3' pipes partially filled, there will be at least 100' of "empty space" inside this chamber. From the prospective of inside the pipelauncher during an acceleration at constant gee, the water inside the chamber stays in roughly in one place with pipe moving up. In order to have a constant gee acceleration the pressure inside the pipe must remains constant, and contray to this there is an ever increasing volume of pipe- BUT "all you need" is for the water to remain at the level needed for the gee acceleration- you need the 60 psi pressing on top of the water- the MOVING hundreds of feet of pipe going by like a freighter train, is *unimportant*- as long as the water "stays put". Therefore "one solution" to the ever increasing volume is to "shut doors" as this train goes up- though having many doors is an added complexity that doesn't "get you much", but having one door (and possibly, maybe two "doors"] doesn't add too much complexity. So this "door" could be about 100' from the top of pipelauncher. Some of this upper 100' section can be used for storage and access for maintenence- and having sealable sections adds another safety feature.