While posting an earlier post regarding pipelaunchers [which btw Nick depleted] I ran across the idea that air escaping from the bottom of pipelauncher at time in which a pipelauncher is traveling fast would *somewhat* or remain stationary.
Meaning if you were in a rubber raft inside a pipelauncher that was accelerating a rocket in a launch; you would be at the same level in ocean depth that is equal to pressure inside pipe needed for this. And this would have some unusual phenomenon associated to it.
Said differently, the air would NOT *seem* to be similar to putting a straw into some water and blowing on the straw- or using an air compressor with 90 psi and doing something similar.
The practical significant of this "principle" is it can in a way extent the length of pipelauncher by having the ocean temporarily act like the walls of pipelauncher.

This principle has little to do with water's surface tension- but rather is more about the time needed for that air pressure which is different than the water pressure to distort the shape of the water. But by looking at some examples that have to do with surface tension, one might have a way of visualizing what I'm talking about.
First, deals with soap bubbles. You can draw a "soap bubble maker" through the air and it can make a somewhat tuber shape and then this shape "collapses" into a more spherical shape of a soap bubble.
Next if one has a pipe in the water, and add enough air pressure, the air bulges at the bottom of the pipe.

Pressure of water is one atm per 33'. A atm is equal to 14.7 psi.
So at 66' you have 2 atm and at 33' you have 1 atm.
If you to create a tube of air which was vertical between 66' to 33' depth and the air pressure was 2 atm. And you made this entire tube of air in say a 1/10th of a second [and this could ONLY be done by something that functions similar to a pipelauncher- you need a pipe that blows air and moves pretty fast] the air pressure at 33' depth has greater different to the ocean pressure, therefore expands faster at 33' level that say the 50' level.

So, you create this tube of air between 66' and 33' and this done in one 1/10th of a second with air pressure of 2 atm. AND the air pressure of the tube of air is maintained at 2 atm.
What happens to this tube of air in the next 1/10th of a second?
The answer is not much- and the bigger the diameter of the tube of air the less significant is any change in diameter of the tube**. The section if tube near the top [33'] will expand faster than lower depth, but at the exact 66' depth you have no expansion all- not in a 1/10th a second or year [if you could *somehow* maintain that pressure and non turbulent environment over any significant amount of time].

**The expansion rate at 33' level will be some quantity of meters per sec per sec. This rate of accelerate will be the same whether the "air tube" is 2" in diameter or 30' in diameter. So taking some high acceleration of even 100 m/s2; in 1/10 of second the expansion in the radius of the "air tube" in a 1/10th second would be .5 meters. 2" diameter tube becomes a little over a meter in size [dramatic apparent change]. Or with the 10 meter diameter, the air tube becoming 11 meters [slight apparent change].