Going back to the u-boat carrier idea a page back, I was for some time wondering how difficult / usefull it could be to launch full fleged UCAVs from a modified submarine.
Take a Ohio or Vanguard and replace the Tridents with a booster launched UCAV carrying it's own weapons (red box) to strike a target, then come back and recover via a parachute (above the engine[yellow] & fuel[blue] compartment), seal the intake & nozzle with an inflatable airbag and be brought back via a crane from the sub in the most aft missile tubes.
Strike potential would be limited, but also quite covert.
Sorry for the long answer, but I tried to give the impression of a profound analyses. Happy reading, I'm looking forward to your answer.
Let's analyze that. We all agree that hullspace is limited and expensive on a submarine and cheaper on a submersible. Guided and ballistic missile submarines would be examples quite close to submarine carriers.
Their tubes for missile storage could be seperate pressure hulls, even external to the whole submarine, but with increased drag and noise. And we Germans have one of the best crazy ideas departements to look up if someone tried to realize it. Well, Allies were afraid so during Operation Teardrop.
If you have the containers as part of the hull, they need full pressure hull capability or a water resistant design that can do outside the pressure hull without damage at any depth and speed the submarine chooses. This makes the whole thing one expensive design with unreliable availability except if you have multiple such platforms with the corresponding costs.
What if you construct something simple to transport the aircrafts under water because if you permanently integrate them into the submarine things get really expensive. See, transporting UAV or some manned aircrafts under water doesn't mean that this payload is as load as the crew of the naval vessel and so requires less precaution against sound emission (quite expensive). Having just a suitable container means also reduced cost for readiness because all big enough submarines can tow it if one has engine problems and even surface vessels can help. This is an idea reminescent of the "torpedoes" used in drug smuggling, quite a cheap asset and it doesn't need the same dive depth capability as the submarine (that saves lots of money).
So my suggested container with aircrafts on board needs to launch them. Here comes problem number one: The wings take up lots of space (that's why I wanted the cheapest subsurface space possible) and the hot fumes of one craft can damage the next or you launch it subsurface (DARPA currently tests a suitable submersible aircraft) or you wait forever (in modern battletime concepts) until the heat is bearable. Furthermore all thrust engines work at their worst the slower the vehicles they drive currently move, starting is ergo the worst scenario for these. How do you solve these problems? Or do you just plan to create a cruise missile that can be tracked back to the firing platform in order to destroy it?
So I'd ask myself what for is such a submarine carrier good for that can't be done better with another platform? It's a stealth carrier for bombing attacks on unsuspecting targets. So let's find some targets.
What about merchant shipping? You could tow such a subsurface aerial asset behind many different ships without giving your armament away to enemy surveillance and range and stealth would allow you to operate this asset with much reduced risk of someone putting 1 and 1 together that you are a cruiser disguised as a merchant. Things get even better if you integrate armour and space for a boarding crew, so you can switch ships and even use the enemy's confirmed own ships for your own ends. So I see lots of potential for an aerial subsurface transported asset in this context.
Other targets with a real submarine transporting your aerial assets would be the things super-stealth bombers were used to take out. Well, I don't believe that the submarine solution is cheaper than aerial refuelling en route.
Next idea, the arsenal ship under water. Redundant with guided missile submarines, requires stealth also during return flight because the spotted submarine platform has a bad self-defense (it was tried to no avail to improve that with the Milchkuh design) or it carries many more additional anti-air, anti-missile missiles and close in weapon systems for self defense. Does this leave any payload or expenditure benefit from the returning aircraft for improved platform performance?
We had this boarding party from the Q-ship, what if you airlift your special forces from a commando submarine? This asset could greatly help retrieve pilots in enemy territory by a quick reaction force. Move down to the autogyro, I think that would be a design that can be quickly attached to all subamrines without great costs.
On a different trajectory than creating a submarine carrier are German developments to integrate an autogyro(WWII) and gliders(current) for observation. Theoretically the autogyro is similar to the currently employed naval camcopter with the add on that camcopter has a pointy object that can fly fast, called a missile (more of an arrow, but well). So a submarine might carry some surface capability close to a frigate with ease by using either the small camcopter or a larger piloted skeleton autogyro with removeable pressure sensitive parts for storage in the pressure hull (I think of a kind of killer egg).
