Sorting doesn't hurt you but it does take time.
Sorting on poor information absolutely can hurt.
If it's suggested that sorting would matter more at 1kyd, then some of this could be due to BC variance.
Well, a bullet's drag is the summation of each drag attribute, and no single measure (like BTO, or OAL) helps in this regard. In reality, acting on any single measure could actually concentrate BC variance.
If you separated 500 bullets based on having the same OAL, what did that do to the BC you're going forward with?
Let's say some of them have same OAL because a shorter bearing length is combining with a longer nose or base length. Or the same OAL resulted from a different size meplat.
Then BC is different with those same OAL bullets.
On the flip side, some of the OAL culled bullets, in reality, may have fallen right in-line with the mean BC of the full batch.
You may have been better, overall, to just leave them alone instead of fiddling with bad information.
Matching BC would be a very difficult endeavor.
I have not found a way to do it that I or most reloaders could afford.