Agreed with all posts.
In addition, the aircraft was statically stable in pitch and yaw, but likely not roll. What this means it's like a general aviation pushrod/pulley setup you might see on a Cirrus SR-22. You can see the aft vertical stabilizer added in the form of a ventral fin in some images in flight, as evidence. I recall flight test reports from the pilot stating the aircraft was difficult to control in parts of the flight envelope. There's no canard or horizontal stabilizer, so you'll not trim more than an overall lift coefficient of around 0.6. This makes the aircraft bad for agility as required for a fighter (read high load factors). Also, look at the planform. There's very little overall aspect ratio. This is why the Northrop X-47 variants have the long slender wing panels protruding from the diamond-shaped bodies. If you don't have aspect ratio, you don't have glide ratio and thus poor endurance.
With an autopilot:
The lack of an autopilot makes it bad for transonic speeds (speed as a form of survivability for attack) where the aerodynamic center moves aft. You'll also have difficulty trimming in pitch due to the shovel nose shape that most low observable designers start with before learning they're bad for pitch. All of this points to the need for an autopilot, built on a nicely arranged 6-DoF aerodynamic database, based a wind tunnel test and good flight simulation work - which is within the art of the possible. The low lift planform is not necessarily bad for an attack role such as the GD A-12 planned. I think this is likely the path McDonnell took if it lived on in another life. Their challenges would sit in increasing airflow to, and the efficiency of, the inlet. Probably involved a move to make the inlet look more like the GlobalFlyer as an example - more conventional and less buried. They'd also need more aspect ratio, so you'd see more slender wing panels protrude and more overall size, in the form of a photographically scaled planform (read larger).
Without an autopilot:
I think McDonnell got away with this one aerodynamically for low speeds (i.e. less the 200 keas). But they'd likely not venture into other roles due to some of the reasons I mentioned above: Too slow for attack and jets, too weak in pitch for fighting. If you're limited to low Mach numbers, why use a turbofan? For low observability? OK, but you'll pay the penalty in endurance simply because turbofan engines like higher subsonic mach numbers. Just look at RQ-3A as an example, the endurance sits around 8 hours. Pretty low for an unmanned vehicle that needs to sit aloft for as long as possible. One remaining role that it could fulfill (poorly) would be manned ISR. You could get away with a steady weight payload (ISR gear), to not confuse the aircraft in pitch upon store release. Problem is the endurance is bad, due to slow speeds (turbine engine with an inefficient inlet). Perfect for a pilot - who has limited physiological endurance - as we read about in the AQUATONE project in one of Curtis Peebles books.
All this to say, it may have become a low endurance manned ISR platform or transonic unmanned attack platform.