Who cares if the hole is round if the neck walls are not consistent? Who cares if the hole is concentric to the center line of the case if the geometry is changed as soon as you put it through a die?
Unless you use a perfectly centered expander plug, the "hole" will never be concentric if the walls are not consistent, and even then the difference in the wall thicknesses will cause the brass to spring back inconsistently, so the hole will no longer be concentric to the axis of the brass anyway, let alone "round"!
And if you use bushing dies without turning, the outside may be round, but the inside will be oval and off center.
IMO, if you don't neck turn, practical concentricity will not be achievable. A reamer will make a nice round hole, but it won't stay round if you don't also turn. OTOH, turning alone and passed through a proper die will achieve the desired results of concentricty AND even neck tension (assuming proper annealing and quality brass).
Brass is turned to a reference point, i.e. the mandrel. Reaming* is usually self-"centering" to the geometry of the hole as it exists while reaming.
*A reamer previously mentioned (Wilson, by pjparker) does seem to be referenced to the outside of the neck. That is the only type of reaming set up that makes any sense to me.