I speak plainly here.
If you suck, it doesn't matter what data you collect. I'm not saying anyone in particular is a bad shooter, but in real life in my experience; when people meet up for shoots, competition, etc face to face….people aren't as good as they seem on the internet. There's more to it than being able to get a tight group every once in a while at 100y. We can roll this all up into a singular category: fundementals.
Secondly, people get lost in the numbers. Numbers matter, but how they matter and to what degree has a little more nuisance, which in my opinion, can be better unearthed reading Bryan Litz's applied ballistics series if you have no formal training. Or even for those that do.
That said if one insist on running BC's over custom curves for instances, you need a few profiles for various ranges at a minimum. This is the correct way to do it. The closer to subsonic you get, the more drastic changes in your data, the more you will need to verify bullet stability, and the more trued the data will usually need to be.
Lastly, as Frank Galli from snipershide says, "The bullet doesn't lie, believe the bullet." Doesn't matter if you have a $600 labradar. If you're .3 mils high, you're .3 mils high. Going back to my first point, that's assuming you don't suck, and it's a reproducible shot, day to day, week to week. In the example of being .3 mils high, you true to that. Depending what range is in question, if you supersonic, you'll want to true velocity first a foremost. Transonic, and subsonic will be more (not exclusively) BC dependent, and going back to my previous point, will require more profiles if using a BC over a CDM specific to that bullet.