But that only works when you can be sure no one's able to go after the battleship with torpedoes or bombs. In any environment contested by a remotely modern opponent, battleships aren't survivable enough to matter.

Sure, but the modern battlegroup concept with have a Destroyer or two for air and submarine defence. The carrier USS Carl Vinson is in the South China Sea right now with 8 Destroyers as escort.

Meanwhile, any opponent we're likely to encounter there has supersonic anti-shipping missiles, against which only CIWS point-defense cannons are likely to be useful. CSG-1 mounts only sixteen such weapons, and each only has magazine space for, at best, 20 seconds' sustained fire before requiring downtime for a manual reload.

While the Phalanx CIWS is inarguably impressive in specification and testing, the thing about the US military is that, whenever we get into a fight, we're always 150% ready - for the last war. It takes us a while to figure out what we're doing in the current one, and it costs us. I see no reason to imagine this pattern shouldn't hold in a new war against an opponent equipped with Moskits.

I thought that the Phalanx was in the process of being phased out in favor of some short range missile system (RAM?)? Phalanx is awesome for taking out subsonic missiles flying a straight trajectory. A mach 3 monster doing evasive maneuvers, not so much..

Think of the problem:

- At Mach 3, the missile is flying at roughly the same speed as the unguided bullets the thing is firing. If it's additionally doing evasive maneuvers it's going to be pretty hard to hit.. (though I guess at Mach 3 it's not going to do very much acrobatics, but perhaps even a little bit is enough to make the CIWS miss?)

- Mach 3 is close to a km per second, and I guess the realistic range for a 20 mm cannon isn't that much, lets say 4 km for the sake of argument. And below, say, 1 km it probably doesn't matter if you keep firing at it or not, it's likely to hit you out of sheer inertia even if you hit it. So that gives you around 3s time to fire and adjust the aim. Not very much..

Sources I found suggest a speed more like Mach 2 when near the surface, and that evasion ends a few kilometers out from the target, well outside the Phalanx's firing range. I'm willing to credit CSG-1's existing CIWS with being able to knock down a Moskit or two - but not twenty, or thirty, or fifty, or a hundred, coming in all at once and, given their range and wide launcher compatibility, quite possibly on more than one axis. That's how we lose a carrier group.

I have seen some mentions of a move toward missile-based defense, which would probably be a more effective option, but most of the same considerations apply - each missile might be much more likely than an individual CIWS round to score a kill, but many fewer missiles can be carried, and the launchers still need time to reload once they exhaust their magazines.

This is also, incidentally, the real reason why DEWs and hypervelocity kinetic weapons - lasers and railguns - are such a matter of interest in the naval research establishment at the moment. They're justified as being a more cost-effective means of dealing with low-level threats, but this is America - the last time our military services had reasonably to worry about being short of funds was no more recent than the mid-1930s, and nobody since has given any meaningful fraction of a damn about the cost per shot of naval weapons - and frugality would militate against the very considerable R&D cost of new systems without new capabilities, in any case.

A much more plausible reason why these new systems receive such interest is because they might eventually offer a solution to the currently insoluble problem of how to defend against a mass attack with supersonic AShMs. You don't see that discussed much in public, because no one with a stake wants to suggest that our enormously expensive carriers no longer offer the power and prestige that merits the cost of keeping them active, but nothing else makes as much sense.