To me the correct answer to all three is, it depends.
First, it's a math problem. Sources of error that affect group size don't add up linearly. They add up the way standard deviations do, as the square root of the some of the squares. Each source of error adds a certain amount of area to the group, and the bigger the group gets, the smaller the change in radius is needed to eliminate that same area, so the less significant it is to the diameter of the group.
YAWWWWWN! (There he goes again. Why should I care about the math?) Well, in this case its because it tells you why some accuracy loading steps matter to some guns and shooters and don't seem to help others at all.
Suppose I turn case necks on a benchrest gun and it improves my 100 yard groups from 1/4 in circles to 1/8 in circles. A 50% improvement! That's the difference between winning and not even placing in some benchrest matches. A very big deal! So, now I load the same ammo into a hunting rifle that normally shoots 2 in groups without neck turning. How much will neck turning improve the groups, assuming it has the same amount of effect? 1/8 in? Nope. Changing a 1/4 in group to 1/8 in removes 0.037 in² of area from a group that started with 0.049 in² of area. A big percentage. But if I subtract 0.037 in² from a round 2 in group (3.14 in² of area), The same change is a much smaller percentage. It will only improve a 2 in group by about 12 thousandths of an inch. Pretty impossible to see, given that it's less than the random group size variation from one group to the next.
So the trick is to address the dominant issues in your gun. What they are will vary with the platform. Are you going to worry about getting bullets perfectly straight into the cartridge, reducing it's runout? Well, Harold Vaughn showed 0.004" of bullet tilt only affected group radius by about 0.36 moa in a tight benchrest platform. But A. A. Abbatiello showed that in a much looser miiltary type match rifle chamber that same 0.004" of tilt made almost 2.0 moa of difference on target. So does it matter to your near 2 moa hunting rifle? It might be the main accuracy issue you have. Or with your bullet and chamber it might be another almost invisible factor. There's only one way to find out.
Components are different, too. Hatcher described loading two stick powders for National Match ammunition one year, both about like modern IMR4320, but one with a short grain and one with a long grain. The Frankford Arsenal loading equipment could dispense the short grain to ±0.3 grain precision (0.6 grain span). It could only dispense the long grain to ±0.85 grain precision (1.7 grain span). Yet, the coarse grain ammunition shot consistently better groups than the short grain loads and was selected for that year's national match load and several records were broken with it. Powder combustion is complicated and is affected by space between grains. My guess is that powder in the particular charge weight chosen behaved such that when the grains were more tightly packed it slowed the flame front passage just enough to make the powder behave as a slower burning powder would. The result was that increases in charge weight were compensated for by decreases in burn rate, resulting in bullet barrel time remaining about the same.
But who knows. A. A. Abbatiello used NM ammunition sorted with a runout gauge. If he'd been able to change the seating depth, would the runout still have been as critical? I don't know.
Twice in my life I've encountered guns with barrels so badly made that the rifling was unevenly deep on opposite sides of the bore. Both of them tumbled and keyholed anything you shot through them from about 25 feet on. One was a S&W m. 41 .22 Rimfire target pistol. The other was a 4" barrel on a Dan Wesson v.15. I've not seen it in a rifle, but see no reason it couldn't happen there, too. It's one instance in which the barrel has to go or no loading practice you have in mind will make any difference to it.
I've also watched fellows so nervous they couldn't stay on paper from prone position at a local 100 yard reduced range match. It didn't matter for them how good or bad the guns or the ammo were. On the other hand I've been pleasantly surprised by how many beginning match shooters will arrive on the line with something looser than Fibber McGee's closet, declaring it shoots better than they do, only to find their best scores improve 10 or 15 points if you persuade them to borrow an accurized loaner rifle.
So, the bottom line is that shooting is a shooter, gun, ammo system. You have to decide where the weakest link is. If you can eliminate the shooter by shooting from a bench, that's one variable down, though most I see can't eliminate him as completely as they might believe by that course of action. Still, you have to figure out what matters to your gun, then address those factors in your loading and not others.
I'll mention one last thing. One year I tried shooting 2520 in my M1A. That gun normally shot about 0.7 moa from prone, but with 2520 it was more like 1.2 moa. I tried adjusting the load, the seating depth and other physical factors. At the time I was using Federal 210M primers, being blissfully unaware of slamfire talk and also didn't know going to a magnum primer could help spherical propellant ignition. What I tried was deburring the case flash holes. Bingo. Groups dropped to the usual 0.7 moa. That deburring never made a hint of an improvement with any stick powder I had in that gun. But with that primer and that powder at the charge weight I had under the bullet I was using in that rifle, it mattered.
So, which loading steps are most important? It depends on where in the particular system the dominant error is located. You can see how, in the last example, if I'd never tried anything but my favorite stick powders, I might have concluded that deburring flash holes makes no difference at all. It's easy to fool yourself into concluding things that are true for your particular shooting system(s) applies to all others as well, but that doesn't make it so. The only thing I know to do is try things until you find out what matters to the weapon(s) in question. Also, based on the math, things that didn't matter when your groups were big can begin to matter as your groups get small.
So like I said; it depends.
Gunsite Orange Hat Family Member
CMP Certified GSM Master Instructor
NRA Certified Rifle Instructor
NRA Patron Member