I don't know what you man by "documented," but it is all anecdotal. From any one individual, the sample size is too small. As with almost everything in Benchrest, there have been no controlled studies with statistically significant data collected and plotted.
On the surface, your question makes sense, but given our sport -- what, 3-4 thousand total participants, including point blank group & score, and long range, it's almost as bad as a "what's best?" question.
Maybe a bit of history would help, though. In the bad old days, case wall variation could be quite large. Guys would chamber a round, and if they extracted it, there would sometimes be rifling marks on one side of the bullet only. Humm, what to do? Turned out, of course, that if you turned the case necks, wall thinness was even, so the bullet was more closely aligned to the bore center when chambered. (The rest of the case, and the chamber itself, still had some influence.)
OK, with some of the brass we have today, case wall thickness is quite well controlled from the manufacturer. Sadly, it isn't as simple as Manufacturer A versus Manufacturer B for all cases; the particular case still needs to be considered. But for .220 Russian, .308, and 6mmBR, Lapua does a pretty good job.
So here's one man's bottom line: It you have a tight chamber (not SAAMI, but cut with a reamer designed to fit a particular brand/lot of cases), and historically, neck wall variation from a manufacturer runs .002 or less, you don't need to turn necks, save perhaps for point-blank BR, where even .005 inch shot dispersement can matter *significantly.*
Alternatively, you can buy cheaper but equally tough brass -- such as Winchester -- cull the cases that are too far outside the tolerance you set, and be at roughly the same place.
When all is said and done, the surety of having equal case neck wall thickness is just simpler to achieve by turning the necks. To get the same effect by culling cases would involve a lot of measuring, and a somewhat higher brass loss due only to culling.