On 20 Nov 2014
US Makes more Friends,
You're talking of current drones. I'm talking of what can be done. They can take more G's because they are not limited by what a pilot can take.
It can be done - with tradeoffs.
The Current Drones are frail by design - the airframe, not the pilot - can't handle >2G maneuvers. This is to keep them light and fluffy so they can loiter for long periods of time.
Bulk 'em up, and you're losing the loiter time and range. And you might need better engines - or less payload, including avionics and sensors. Want to add more capacity to operate independently? Then you'll need to increase the weight of everything, the spread of the wings (less stealthy now, too), and of course costs.
Will the new drones still be C-17 transportable at the new size and weight?
When the costs sharply increase because you're turning them into independent, computerized combat aircraft, can you afford to swarm them and lose them?
If you don't have vision, you at least have radar signals from the missile guiding system.
Assuming the missile is active radar guided. It could be passively homing in on the drone's radar. It could be an IR guided missile. If the drone is equipped with ETDS, that's another layer of complexity, plus you'd have to give it the agility to actually jink and turn inside a missile on it's own. Not to mention the current drone turboprops can't really reach speeds that making turning inside missiles easy.
AAA Gunfire would also be a problem (gee, I wonder why drones are built to fly at high altitudes?). A human pilot can see the tracers. A dedicated ground attack craft has better armor and more countermeasures.
These problems are being solved. In a few years these will be commodities. I would think the military would have top notch solutions (for what we pay for it).
We've known about V/STOL for a while, but making a plane that can toe the line with similar generation fighters while being safe to operate has still not been achieved. Check out the operational accident rate per hours flown with the OV-22 and AV-8B.
Land-based drones are similarly limited. The US Army was testing one that was louder than the loudest weed-whacker I ever heard. Good luck on a platoon sneaking up on the bad guys with a drone that sounds like a whole band of Hell's Angels on Harleys riding through town.
A good drone can act quasi autonomously until the kill decision. You don't need that much communication. Or you can relay it through many different drones.
Again, more coding, more capabilities, increased cost. Communications are active emissions that can be detected, jammed, and attacked with passive homing guidance systems.
You'll get to a point where an autonomous drone is just as, or more expensive, but still less versatile in the face of a diverse-threat environment that a combat aircraft with a crew.
What if you have hundreds of drones made of plastic emitting radar waves intermittently flying above the ceiling of enemy planes? There is a point where you just to admit that there will be radar waves and stop worrying about it.
Problem: MQ-9 Reaper
Cost: $17,000,000 in 2013 dollars
Cost: $550,000 in 2010 dollars. That warhead is overkill.
SA-17 aka BUK
(and many other SAM models)
R-27, R-77 and others.
Problem: Needs satellite, at least in "kill phase"
Not saying it isn't solvable, but the more complex these things become, the more the cost savings evaporates and the more fragile the system is overall.
They become too expensive to swarm the enemy.