In a firearm, specifically. This is an interinting topic to search, as many people have opinions on it, but few have any hard data.
I have a varmint rifle that will shoot <1MOA all day long, and I guess I've gotten spoiled by it. I also have a Hi-Point carbine that shoots about 3 MOA. For this gun, that's probably not unreasonable, but why be reasonable if it presents a chance to tinker.
I found a thread in the Hi Point discussion group that mentioned that the barrels of H-Ps were not crowned. The moderator pointed that crowning cost money, and H-Ps were built to a price point, not an accuracy standard. This seems reasonable, as the easiest way to get a sum-MOA gun is to put about $3000 on the table, and wait for it to get built. Still, I know that some massaging will improve accuracy, and it doesn't have to be prohibitively expensive, so I looked into crowning.
For you non-gunnies, the crown is the recess cut into the business end of the barrel of the gun. The form varies, depending on how fancy you want to make it. Most guns use a double radius, so the end looks like a donut, radiused into the hole, and out to the outside diameter. This is easy to cut, and not expensive. There was a discussion at Snipers Hide on the relative merits of crowning that was reminiscent of .45 vs 9mm arguments. There was one post there that cited some actual data, suggesting that the difference between a nice square crown, and a 30 degree hacksawed barrel was mostly to move the shot group, although it was noted that if you cut off your barrel with a chainsaw, you can expect your groups to increase from 1 MOA to 1.5 MOA. One could conclude that the primary function of the crown is to prevent damage to the inside of the barrel while the gun is banging around behind the seat of your pickup truck.
Bill Johnson has a more exhaustive piece on accuratizing that has the advantage of having reasonable extensive documentation, and the inclusion of a control rifle. His test rifle started out at 4.5MOA which is worse than mine, so I felt this was a relevant study.
Polishing the barrel using Dave Tubb bullets with abrasive embedded into the jackets took him down to 2.67 MOA which is an impressive drop for an easy job, plus the price is reasonable, and you get to go shooting in the process. Shooting abrasive bullets down your barrel sounds scary to me, but this seems to work.
A trigger job made more difference than I would have expected, but hey, whatever works. I suspect that most trigger jobs involve simply polishing the parts to get rid of the scratchyness and reduce the friction. To that end in might be fun to see what throwing a disassembled trigger group into a tumbler with metal polishing media would accomplish.
Bedding the action helped, but only on the looser of the two rifles. Once you get to a certain point, the return on your investment falls off rapidly. It helped for Mr. Johnson, whose 4.4 MOA gun now shoots <1MOA. I'm not sure what bedding the action might entail on a gun that the furniture comes in a left half and a right half held together by screws that pinch the halves together over the metallic innards, but I guess I could have a look while I'm looking into the trigger group.
Keeps me occupied during the winter, I guess.