Here's you're original Question...
If you are getting +/-.0005 (i.e. w/in a .001 spread), run em. Even if your measurements are not perfect.
Spife gave you the same answer, with data.
So did SteelHead - you're overthinking it.
You' said " About every 8-10th rd I would have a casehead to ogive varaiance of .003 to .006 ". Is this perhaps a compressed load, or close to it? I would check to make sure you have consistent neck tension (likely not the problem), and that you have a seater stem that works well and consistently with the bullet you are seating. This is more likely a seating stem/pressure/neck tension issue than a bullet variance problem.
Just thinking and trying to help.