Ok Ok,
The last 10 years refers to the volume of hours and experience we have gained on these aircraft consistently on combat operations and the improvements in capability/reliability is becoming exponential due to the sheer volume of investment from the US. And that this technology fits the current battlespace.
Secondly is this debate about autonomous or unmanned I believe the US Army is going for combination as they are two very different concepts.
And yes there is definitley scope for autonomy in the battlespace (the processing power for certain applications already exists), CAS and ISR id targets is only a small element which at this time require human interface. Barnstormer, Your daily mail esque reference to being on the ground and being fired at by something which hasn't reliably/positively Id'd the area well is a bit odd, ask what happens when MGLRS/ arty is fired at range, the firing unit (which could be a computer) rely on the the accurate input of the Jtac/troops likewise any aircraft/pilot requires the same service, I don't think anyone would suggest a computer could positively ID enemy forces in a danger close situation. But why can't a UAV fly autonomously over the battlefield and accept fire orders in an appropriate format direct from the ground troops? And deconflict as we currently do. As an aside nothing is truly autonomous or without supervision in the battlefield or off for that matter, so a decision matrix could easily be programmed to know when to ask for human input.
In addition, autonomous lethal weapons already exist (goalkeeper/phalanx) which are fully automatic and will shoot anything down that has an appropriate speed/height profile. So why can an airborne platform in the maritime/air/land environment do the same when the tactical situation allows, ie hot killboxs/ no fly zones.
The reference to airbus fcs illustrates that software is now so reliable and redundant and capable of complex logic that people are happy to stake their lives on it, on a daily basis, as for the pilot being a supervisor, definitley but as a pilot I know my accuracy of flying and nav skills will be beaten hands down by a computer, and in reality when does a pilot override the system and succeed vs how often does the pilot do nothing because the system works just fine thankyou. I was trying to illustrate reliability.
I still don't think UAV RPA etc are the be all and end all but not to think a bit broader on this subject would be doing the future of aviaition/warfare any favours.
Finally I was only submitting an opinion and never claimed to know so much about this subject, however I did go to stores to try and sign out a high horse but they were all out, I think vec got there first. I am still learning.