The training requirements for catapault launch/arrested recovery ops are such that the aircraft procured would not be available to do anything else. They would be an asset which would effectively be tied to the carriers.
The theory is that a STOVL type will be able to conduct both carrier ops and land based, and will thus be more flexible and more useful, more often.
The implications of command over carrier-capable assets was discussed recently on another thread.
So in answer to the original question, the answer is B) we are getting carriers so we'd better get some JSF.
As to what
relevant capabilities we will get that are not already available using other platforms, there are very few.
Like the OP I'm not convinced that the carrier argument is fully valid. I can see that there are theoretical scenarios in which carriers would be the only option (you only have to look back 26 years to see the last example.....
).
Unfortunately, though, there are equally compelling scenarios that could be used to support the acquisition of similarly high-cost and seldom-needed capabilities - from the ability to air-drop paras in brigade strength, through the ability to mount another Desert Storm/Granby scale effort, to the ability to mount a large scale armoured offensive, or even the use of strategic heavy bombers.
We can't afford any of these - nor, according to the defence assumptions, can we afford to do the kind of large scale autonomous ops to which many of us would aspire. In that strategic and budgetary context, CVF is not affordable, and should not be acquired. We are not the major world player that we once were (and should still be). CVF and JSF together do represent a MASSIVE amount of money, that would be better spent on the capabilities we actually need every time out - not just helos and transports, but tankers, SEAD, recce and land based FJs as well.