I do not understand why someone would "true" their velocity, if in fact, i know exactly what my velocity was during the given test.
Chronographs, just like scales, can have a margin of error.
This said, there are other factors in play when verifying data/dope. The basics in particular order: all inputs are absolutely correct (this includes scope height, zero distance and zero height), exact distance to the target is known and scope has been tested to insure mechanical deviations in the turrets are considered and a correction factor entered. The checking of the scope's ability to track to the exact value of the turrets can be done using a tall target test, although there are jigs that one can use to accomplish the same thing. When doing a tall target test, I do mine at 100 yards, the distance from the turrets to the target MUST BE EXACT. If you are off a yard, the information will be tainted and be "just another bad input".
After I develop a load and have decided on a charge weight, I will do a tall target test. This consists of 9-10 shots over a chronograph while doing the test. This lets me check, or double check, the scope's tracking and obtain my MV at the same time. I've been fortunate that all of my scopes have tracked well but I've had students with some that were pathetic and had to be sent back to the factory for calibration. Some amount of error can be compensated for in your ballistic program but if grossly off, send the scope in and have it checked.
After getting a MV and checking the scope, I head to the range. As mentioned above, many shooters, myself included, true their MV at 600-800 yards (it needs to be a distance where the bullet is STILL supersonic) and then true their BC at 1000 yards or farther. BCs change as velocity changes and is more critical farther downrange as the bullet slows down. Guys that are really into checking their BCs sometimes use the Shot Marker system downrange (1200-1500 yards) and get their MV at the target and adjust their BC accordingly. Most of use don't have that luxury so we rely on POI on the target which usually works well enough.
REMEMBER, the target (or bullet) doesn't lie.