It's the thing with fins, a GPS and a little actuators to move the fins, attached to a projectile isn't it?
If they can make drones for a few hundred dollars, it isn't the true cost of the projectile, its the price the company thinks the military would value it at. They guessed wrong it seems.
I sort of guess, the company figured that estimating the price at 80k, the military would buy the guns, and then get stiffed with the 800k price tagged once they were locked in, rather than admit they'd been screwed over.
Clearly you may aswell buy a missile and get better accuracy, range and more launch options than this. So the munitions supply priced it wrong and the military decided to cut their losses on the project.
The government has a very keen insight into how much it costs to produce these projectiles. This isn't a car lot where they just show you some bullshit invoice. Even the government doesn't think the contractor is over-charging, just that the cost dramatically exceeds the original estimate. FWIW the guns and projectiles are built by different companies. This is a classic cost spiral. When the ships cost too much they reduce the quantity, making each copy cost more to build. When you only have three ships you reduce the quantity of the projectiles and will not ever receive the economy of scale. For example, Small Diameter Bomb II: Initial production run unit cost was double what the latest unit cost is in production. That improvement happened in a couple of years with an order of ~3000 units. In the original AGS plan, a cheaper, non-rocket propelled, round was also to be developed for the gun, that was canceled early in the program. Don't know why.
A lot of people like to think cost over-runs are just slimy contractors grabbing for the brass ring. Sometimes, but it is usually much more complicated than that. Science projects: both the government and the contractor reaching too far forward for technology that isn't ready for prime time. This isn't just being over-sold by contractors, the government has subject matter experts of their own that assess the state of the art. Requirements creep: When originally conceived a weapons system is designed to do a particular mission and counter a particular set of threats; missions and threats evolve. LCS is a good example. Originally intended by the procurement organization to be a cheap to produce using commercial standards (COTS use where not appropriate is a recurring theme in defense procurement, leading to cost growth later). The contracts for the original two ships for the competition were awarded. Only afterwards did the Navy come back with "the people who actually have to operate the ships in harms way have told us that commercial standards for hull construction are not adequate." Funding profile: When a contractor bids a project, they, along with the government, put together a plan. The contractor has to assume that the post-negotiation contract will be funded per the negotiated plan. Unfortunately, politicians don't have any skin in the game (unless it is being built in their district) and have a habit of diverting funds to cover the latest fire or a deal with an overall decreased budget (remember sequestration?). Obviously this is being penny wise pound foolish because in the long run it cost the contractor a lot more money to slow things down and then ramp them back up. Which leads us to budget proofing: regardless of efficiency, contractors are driven to spread out production across as many congressional districts as possible. Don't blame the contractors here, this is just them responding to pork barrel politics. Which is our own fault for electing representatives, not based on what they will do for the country, but how much money they can bring back to the district, whether the project is needed or not. These are just a few examples of cooperative inefficiency between the DoD and defense contractors.