>Currently the velocity of railguns is roughly equivalent to navel guns
Nope, the 8MJ railgun was hypersonic (I think it's at 32-64MJ now). It sounds much louder than the 5" when it fires. It needs the extra velocity as the shell turns into plasma when it hits a target, eliminating the need for explosive in the shell.
The reason for the shells that no one seems to have pointed on yet is the US Navy also need land-attack capability (The Marines continue, to this day, protest the removal of BB's because they love dropping 16" 1.5 ton shells on fortified beaches, See also, LRLAP). If you look back to the Iran-Iraq war, while missiles were used frequently most things were still finished off with 5" / 75mm. The truth is that those rounds cost much less than a standard missile, and a railgun projectile should cost even less as there's no explosive handling required for it's rounds (either HE or propellant). This is all in addition to the massive improvement in safety the CNO frequently cites (which is also important)
>High endurance aircraft that can strike from extreme range and attack submarines with surface strike capability might be the order of the day. A submersible destroyer for example could get in close with heavy weaponry, fire a salvo, and then dive before enemy systems could target and strike it. Such a thing would be vulnerable to enemy attack submarines but then you could just escort it with a flotilla of attack submarines to act as defense. You could even add some drone carriers. Submersible aircraft carriers were built by the Japanese in WW2. Consider what you could do if you gave such a design a nuclear power plant, expanded the size to Nimitz proportions, and replaced the planes entirely with more compact drones.
While I personally believe we would be better served with elimination of the surface fleet for submersible craft for many reasons (difficult to target, risk reduction to asymmetric threats such as a fishing trawler filled with explosives, reduce need for so many ships in a carrier group, etc.) the "Japanese WWII aircraft carrier" was a sub with a built in storage for three seaplanes that had to be assembled before use and disassembled on retrieval (based on memory of the wikipedia article). The sortie rate and number of aircraft supported of a CVN is much, much higher than this design can support, so we would either need a massive number of them (more $$$ + extra manpower), or need to make a submersible carrier meeting the requirements of a modern CVN. Plus, I would guess if anyone followed those Japanese seaplanes back to base that sub-carrier didn't have a lot of options to save both itself and the aircrew, as it didn't have enough planes to maintain a CAP or enough armaments to enforce a defensive perimeter.
If you Nimitz-sized one, the initial cost would be massive (a typical CVN is over $3 billion currently), and there would be a lot of technical challenges to work out (sealing aircraft elevators against water pressure, for instance). I believe it would be worth the one time high R&D cost in the long run, but it would be a hard sell in current economic times. You also have to think about risk (what's the total cost if one sinks accidentally? How many ways could that happen? Can I evacuate a sub with 4000 people the same as one with about a hundred?