You said it would be bad of it broke up after leaving the barrel? Wouldn't that cause a sort of shotgun effect? Which I guess would be bad if prediction is what you were aiming for but wouldn't that be good for say shooting up a base or large group of fast moving ships?
Two reasons. First, presumably your ship is not fighting alone, but is fighting alongside other ships. This is a bit of a presumption (if your situation is bad enough that you can't afford what are essentially metal slugs, this might be a jury-rigged pile of junk that you can only put into the field in isolation), but it does mean that your engagement envelope just expanded to include any of your escorts that happened to be on either side of the firing line between you and your target. The second is the bane of so many things, the inverse-square law. Basically, as a conical influence moves outward as a linear distance, the area covered expands proportionately to the square of that distance. In the case of a directional radiator such as a laser or our space shotgun, this actual fall-off varies (lasers, f'rex, are much more susceptible to beam divergence due to diffraction), but note that for any cone (that is, neglecting inter-particle collisions that may send your projectiles veering off wildly) with a fixed firing arc
θ, the area of the circular area of effect (the base) is going to be proportional by that square to any linear variation in the distance
h.
r / h = tan θ
A = πr² = π(h tan θ)²
A = π h² tan² θ
So, let's put this into perspective. Say you're one hundred thousand meters away from an object (that's a mere one thousandth of a light-second, very close range). Instead of a solid object firing alone a line (
θ = 0), you accidentally fire this thing off in a shotgun spread of...30°. You've just spread your asteroid across over ten billion square meters of area by the time it reaches your target, even assuming it all reaches its destination in a conical pattern (it won't, by the bye; it'll be a sectored sphere, but I'm getting a bit of a headache at this point). Assuming it's...say, the size of 25143 Itokawa (a respectable asteroid at about half a kilometer on its longest arc chosen mostly for a very convenient order of magnitude for its mass: 3.51*10
10 kg) and that it somehow spreads perfectly evenly (it won't, but here we're into "perfectly spherical cow" territory), you are now delivering approximately 3.31 kg of space dust per square meter. Not completely awful, but remember that (a) you won't be reducing the asteroid to sand as a part of your mishap, but rather randomly moving chunks that can be individually tracked and (b) you presumably needed the bang that came from that magnitude-ten difference, otherwise you wouldn't be firing something that size in the first place.
By the bye, a kinetic space shotgun is arguably going to be the worst possible weapon to deal with highly-maneuverable ships, because they can track and dodge the projectiles much more effectively, and a miss is as good as a mile, whether it's by kilometers or millimeters. It's the reason why bomb-pumped warheads, that so-recent staple of mil-SF that came about as part of some sort of Teller craze, are delivered so close to their targets by missiles before going off. The inverse-square law is to SF physics what the square-cube law is to SF biology; it's a spoilsport.
Finally, though, imagine one final thing by analogy. Imagine that you've created a sniper rifle, only every so often it fires a shotgun blast at random. You can't control this, you can't select for it, and you've basically ended up in a situation where you're going to have your gun randomly firing in one or the other "mode" regardless of which role you place it in. In other words, no matter whether you use it as a sniper or shotgun, you're going to have to deal with misfires. It's not going to do particularly well in either role, in essence.
NOTANEDIT: Also, Gig did this much faster than I. And Amperzand points out the one reason you would use such a weapon - close range against something that can't dodge.