That sounds a lot like the design issue that MegaTraveller came up with -- all those power-hungry weapons and defenses just don't get used 90% of the time, so why power them all of the time with a gigantic central powerplant. Those weapon emplacements could have their own capacitor banks and small power plants, ready for action.
More moving parts to maintain. More reliability testing to perform. Some devices work great if well maintained, yet continually operating in contrast to devices that are rarely used and instead just sit. More components mean more spares and more maintenance.
Redundancy is a value all its own, but a larger motor should be more efficient than several smaller motors. Consider locomotives. These are built as large as they can be. Rather than sticking together a bunch of smaller locomotives, it's better to hook up fewer larger locomotives. I won't say there are no applications, but outside of vehicles chasing speed records, I don't know of any industrial applications where several smaller motors are used rather than fewer larger ones. Large cargo ships use one or two enormous motors to drive the ship rather than a bunch of smaller ones sharing the same shaft.
After the 747, when was the last airplane to get four engines? And how much are the modern two engine planes really using them for redundancy and reliability than out right necessity? C17? Airbus 380? Not saying they don't exist, but it seems more rare than not.
If anyone is on the sharp edge of the bang/buck equation of power plant size to value received, the transportation industry gets my vote. They seem to know what they're doing, and they seem to go big.
If the larger power plant has the resolution and capability to scale up and down to meet demand, then there's lots of reasons to go with the big plants over several smaller ones.
As for power loss, what TL did the room temperature superconductors become commonplace?