stubbicat, you ask good questions.
Here's what happens when a rimless bottleneck case is chambered then fired. Note that this all happens even if the case neck's only .001" smaller than chamber neck.
First, when the bolt's closed on one, the extractor at the back end pushes the case head opposite its contact point on the rim/extractor-groove point. The case head stop against the flange on the bolt head or chamber wall at that point; depending on the dimensions and alignment of each. Typically, the case head's off center relative to the bore-chamber axis at this time. Mauser style ejectors push the case head to one side. Winchester post '64 ejectors in the bottom bolt lug push the chambered round against the top of the chamber.
Second, part 1, if the ejector is a Mauser style, it's not touching the case at all and has no effect whatsoever on positioning the chambered round. The case front to back position could be anywhere in a range equal to head clearance; the difference between chamber headspace and case headspace. If the round's heavy enough to over come what the extractor's pushing it, it may well rest on the bottom of the chamber. Depending on the difference in case body and neck diameters relative to the chamber at those points, either one may be clear of the chamber at those points. The case axis will not be in line with the chamber/bore axis.
Second, part 2, if the ejector is in the bolt face parallel to the firing pin and is spring loaded, that pushes the case forward until it stops against something. Typically, it's the chamber shoulder that stops the case shoulder from going further forward. There's enough clearance from the bolt face to the extractor claw for the case rim to move back and forth a few thousandths. As the case shoulder's moved forward against the chamber shoulder, it centers there as its angle is the same as the shoulder. Doing so will raise the case body off the bottom of the chamber as the case shoulder well centers in the chamber shoulder. The case neck is fixed in position on the case shoulder. It moves with the case shoulder. If it's well centered on the case shoulder, it'll be well centered in the chamber neck at this point.
Third, part 1, the firing pin's released and moves forward at about 5 mph. It strikes the primer in the case head. That moves the case forward in the chamber until it stops against the chamber shoulder. As both case and chamber shoulder are at the same angles, the case shoulder is now well centered in the chamber shoulder. The case neck moves with the case shoulder and now is also centered in the chamber neck; if it was sized straight on the case shoulder in the first place. Like above, the case body moves up off the chamber bottom, but only if it just happened to be there to begin with. There's enough force to set the case shoulder back .001" or more depending on shoulder shape /area and firing pin strength.
Third, part 2, same firing pin speed and striking things happen as above, but the case is already full forward in the chamber from ejector spring forces putting it there. So it gets driven forward an extra .001" or more. And the case neck is centered in the chamber neck as above, too.
Fourth, the firing pin finally dents the primer cup then crushes the primer pellet against its anvil a few microseconds after the case being slammed into the chamber shoulder and the round fires.
Main thing to remember is with cases headspacing on their shoulders, the case neck gets centered in the chamber neck slightly before the round fires. Whatever misalignment it has to the case shoulder is what moves it off the chamber neck axis. If the case neck's only .001" smaller than the chamber neck and the chamber neck's .001" off center from the case shoulder axis, it'll be .001 inch off center in the chamber neck when the round fires.
Note also that the case body is clear of the chamber body all the way around except at its back end at its pressure ring where it's pushed aside by the extractor. As there's no such thing as perfectly round cases and chambers, it helps to keep the case body diameter a thousandth or more smaller than that of the chamber. If there's not enough clearance for those variables to not have a case positioning issue, then accuracy suffers.