Apparently this hypothetical drone you're talking about has the evasion of a missile, and the stealth capabilities of the latest $Billion stealth aircraft, while 10,000 of them put together costs less than a single M1 Abrams tank.
It can also fly at speeds to evade our aimbot CIWS or aimbot CRAM defensive guns and carries enough of an explosive payload to damage an M1 Abrams.
Color me skeptical. I find it unlikely that you can make such a device for $1000ish.
By their nature, a drone could be harder to detect than another threat. A small drone will have a small radar return, and battery powered electric motors will generate less of a thermal signature than a rocket motor.
I'm not at all saying that a CIWS or CRAM would be totally ineffective against a drone or a swarm of drones, but that they may not be the impenetrable shield, either. There are limitations to sensors, target classification, rules of engagement (will the CRAM automatically light up a drone with 50 rounds if the backstop is an office building or residential block?), and ammunition capacity.
Let's take CRAM, AFAIK they are mostly installed to protect against indirect fires, so that may imply that they're tuned/configured against certain incoming projectile trajectory, speed profiles, and radar returns. Bird like radar signatures may be filtered out, or level flight trajectories that do not indicate an impact within the protected area. Nothing that can't be changed, of course, but such change might come after a few bloody engagements.
In a true future war scenario it wouldn't be one hobbyist drone with a jury rigged mortar round against a CRAM installation either, I would expect there to be drones flying around blasting out wide-band EM hash, dropping lines of chaff, and using other measures to clutter the battlespace and confuse and overload defending sensors to allow offensive devices to get through and damage or destroy their targets.
Or maybe, the current state of the art of missiles, is superior to that strategy?
Consider 1 hypersonic missile, 1 supersonic missile, and maybe 10 subsonic missiles, all cruise missiles flying no higher than 100m off the ground.
Since they're so low, they can only be physically detected at a range of 10km (they're literally "behind the horizon" and cannot be detected beforehand).
With proper timing: all 12 targets pop up on the horizon simultaneously. The 10 subsonic missiles will hit the ship in 30-seconds, the supersonic missile will hit the ship in 15-seconds, and the hypersonic missile will hit the ship in 5-seconds.
They are all flying at a randomized flight pattern at ~2G lateral movements. Its not like missiles fly straight at their targets these days, they fly monte-carlo randomly to throw off defense systems.
------------
Now consider the drone swarm flying at 100mph that crossed over the horizon with the 12 cruise missiles. The Drone Swarm will arrive in 2+ minutes. Do your defense systems even care?
Do the drones have any significant amount of lateral G's that they can pull? Can they fly at 2G random flight patterns to throw off air defenses?
In your naval combat scenario, I think you're absolutely correct and the proliferation of autonomous drones will not cause a seismic shift in the battlefield. Unless we go mad scientist and throw in autonomous mobile mines that with a nuclear warhead that lurk around until they home in on the acoustic signature of a CVN's screws from the SCS to Guam. But we already have attack submarines.
----------
I've always imagined autonomous drones as having the most impact in urban combat, or 'traditional' divisional level fights. We've seen the early impact of this technology in Crimea and most lately in the Armenia and Azerbaijan conflict.
Unmanned Predator drones shoot Hellfire missiles at targets.
Stealth RQ-170 Drones stalk the enemy and obtain intelligence without tipping off the enemy.
Maybe other drones exist (maybe so stealthy that their existence is still unknown to the public), but we already know of these two strategies from public reports in Afghanistan / Iraq.
--------
We also have models for what we expect the next 10 years or 20 years of (potential) war to look like. A likely scenario is that China / Taiwan relations sour dramatically, and China decides to attack Taiwan for some reason, and the US rushes to Taiwan's aid.
Under such a scenario, the USA would be armed with whatever our Carrier Strike Groups can carry (CIWS, Cruise Missiles, a myriad of supporting destroyers and submarines, multiple air-wings, etc. etc.). While China would be armed with its huge Air Force, its increasingly huge Navy (of mostly small ships, but a few smaller carriers are on the way), and most worryingly: China's high production output can build huge numbers of cruise missiles.
----------
There's probably a few other locations in the world where USA vs Russia could get into a ground conflict. But... lets take the China vs Taiwan scenario for a sec and think.
Where does drone warfare fit? Stealth Drones for detecting enemy movements would be key: both for US and China. Carriers are fast and expensive: China absolutely would want to neutralize them. But how does China find a group of sea-vessels traveling at 40-knots in the first place?
And that's assuming that those drones aren't shot out of the air by anti-air defenses.
---------
Maybe the detection problem is solved, and now China is ready to fire upon a US Strike Group (not necessarily at the carrier, but maybe at its anti-air destroyers / cruisers in preparation for future strikes).
It seems like the obvious weapon of choice would be China's hypersonic cruise missiles: traveling at Mach 5 and with coordinated swarming AI, its the obvious way to avoid CIWS / CRAM / Patriot missile defenses.
Once the anti-air defenses are down, China can continue its attack with cheaper subsonic missiles to defeat the rest of the fleet. No drones required, aside from maybe the initial recon missions.
-------------
As far as people can tell: it seems like it will be a game of cat-and-mouse. If the US Carrier Strike group can avoid detection, the US would win the combat.
But if the US Carrier Strike group is detected: then its all over. China has more than enough cruise missiles to win.
--------
I'm not seeing how, or where, a swarm of cheap electric-motor autonomous quadcopters comes into this playbook. That's bad for stealth / recon, because the enemy knows you're watching them. (AKA: The Aircraft Carrier will just launch an F-22 and shoot down the swarm, blinding the would-be intelligence). To be successful at its recon mission, the drones have to secretly watch and pinpoint the carrier strike group, or survive a dedicated attack from the Aircraft Carrier's myriad of squadrons at its disposal.
If we're already past the detection phase for some reason, then there's no need to use drones. I think the expectation is that China's Mach 5+ hypersonic cruise missiles will instant-win the combat. Those move too fast to be shot down reliably. (Patriot missiles only fly at Mach 4)
