If the point about China/Russia increasing their relative punching power vs. the US doesn't feel significantly to you, what about when they are sold on the open market to any country that wants one? Any country in the world being able to take out a carrier would change the way that power is projected in a very big way. Suddenly attack and defense are much more asymmetrical, which favours the small actors (this would work against Russia and China too, but they have less to lose than the US does in this regard).
On locating a carrier:
"Few if any nations today possess an assured capacity to track carriers continuously. All of the relevant methods — radar, electronic eavesdropping, electro-optical and acoustic sensors — have major drawbacks such as high cost, vulnerability to preemption, and inability to precisely discriminate. While that may change over time, aggressors will still face a daunting task in penetrating the layered defenses of a carrier battle group."
"The areas in which carriers typically operate are so vast that adversaries would be hard-pressed to find them even in the absence of active countermeasures by the battle group... The most practical way to conduct surveillance over such expanses is with satellites, which have a much wider field of vision than sensors operating inside the atmosphere. However, China does not presently possess military reconnaissance satellites capable of finding a carrier, and the commercial satellites to which it might turn take days or weeks to task and deliver imagery -- making them useless for finding a continuously moving vessel."
On tracking it:
"In order to keep up with the carrier's movements, an attacker must establish a continuous track of the vessel using some combination of land-based, sea-based, space-based and airborne sensors. Moreover, the track must be sufficiently precise so that it can provide targeting coordinates to weapons when they arrive in the carrier's vicinity. As of today, even the United States has difficulty accomplishing such a feat, and no other nation is close to having the requisite capabilities."
"Space-based or airborne radar with the capacity to track moving ground targets (such as the Joint Surveillance and Target Attack Radar System, or JSTARS) has the greatest potential for providing reliable targeting information, because it combines precision with a capacity to operate day or night in all kinds of weather. However, few prospective adversaries are likely to have such systems before the end of the next decade."
Summary: Note that this article is circa 2001, so when they say "few prospective adversaries are likely to have such systems before the end of the next decade", I read that China and Russia could potentially have them now.
So I think there's a strong case for carriers being hard to track historically, but it sounds feasible that large nation-states have gained the ability to do so in the last 5-10 years, and it seems likely that that capability will be democratized further by low-cost commercial imaging (E.g. see SkyBox for the state-of-the-art in civilian high-resolution low-delay satellite imaging).
That's a technique from WWI. For a modern long-distance radar a carrier is like the Sun in the sky (as you would expect from a huge pile of metal on a plain surface). It's too big, and also too "loud" in terms of radio emissions to be stealth.
In short, over-the-horizon radars are not exactly precise. AFAIK no OTH radar to date can be used to provide a firing solution, they all are warning/detection radars. They can also be jammed or destroyed, as they are huge and can't be moved or hidden easily. Finally, it's not enough to provide an initial targeting, you also need to guide the missile during its flight as carrier moves and evades. This is a huge problem, given that any signal can and will be jammed.
I think you underestimate how silent the carrier can be.
[0]: http://lexingtoninstitute.org/wp-content/uploads/aircraft-ca...
What about optical satellite observation?
Also, in wartime there will probably be decoys and other counter-measures that might be hard to differentiate from satellite resolution. You would want accurate tracking for your $50-100m launch that may or may not start WWIII.
I find these two assertions very surpsising. Which part of the process do you expect to be slow and what do you mean by slow? Why do you expect accuracy much smaller than the size of a pixel in an image?
The obvious solution would be to just have many megatons of kinectic energy hit near the carrier group via a decently sized kinetic, but the US would treat that exactly as a nuclear attack by a foreign power and respond in kind. It would probably look like one to central command as well, mushroom cloud and all.
I think the idea that these things will be stealthy, maneuverable, unstoppable, accurate, and undetectable is a little much. These weapons come with quite a bit of liability and might be impractical in any non-total war scenarios. The US is not going to let China and Russia blow up carrier groups with multi-megaton weapons and not respond in kind with its own megaton weapons, either nukes or large kinetics against military bases or capital cities. Once that hotpoint is reached, nuclear war is assumed. A kinetic strike against USS Ronald Reagan will mean a kinetic or nuke strike against Beijing or Moscow. The Chinese and Russians would know our escalation policies and would act accordingly if they don't plan to start WWIII with these weapons. Whether these policies have been officially written yet is the only real question here. Its safe to assume multi-megaton attacks against the US will absolutely not be tolerated and will not be treated with a conventional response.