Launching systems aren't modular add-ons. Steam systems require piping, accumulators, steam generators, actuators, as well as the piping from the heat source (reactor). EMALS is just as complicated, with the electrical wiring, switches, busses, not to mention the catapults themselves. Gerald R. Ford already is at sea doing pre-commissioning trials, and John F. Kennedy is scheduled to be launched next year. To go back to steam systems would require them to be installed on Enterprise, which would mean a redesign of the ship, involving delays and cost over-runs.
Now, why EMALS instead of steam catapults? Efficiency. Steam catapults are heat engines, which use expanding gas to do the work. The maximum efficency of an ideal heat engine is only around 30% to 35%. Also, steam cats apply maximum force at the start, and fade out at the end of the run. EM systems have much higher efficiencies (80-95%) because there is much less energy loss to heat. Also, the force can be controlled and programmed for the length of the shot, so it minimizes stresses on the airframes.
Yes, there are development costs, but in the long run money is saved in energy efficiency, less wear and tear on the airframes, and lower maintenance costs. The technology works, as seen in the shore testing. There are always glitches in all systems in sea trials; that's why we have them.
