This is a real phenomenon but it may affect you differently than you'd expect.
1) Groups will be their smallest, all else equal, if the shooter maintains their cheek weld between each shot. This ensures that any parallax error is held relatively constant through the string.
2) Groups will best reflect the true center of the cone, all else equal, if the shooter breaks position between each shot. This randomizes any parallax error which may occur on each shot, though it can't eliminate inherent bias in your cheek weld and shot process.
Ideally, get behind the gun while it's held steady at the front and rear. Without touching the gun, move your head behind the scope to observe the reticle moving relative to a target. Adjust the parallax knob and repeat, until you can no longer observe appreciable parallax error in this test. When you've done this, note the setting - it will likely not line up with the settings on the knob.
Of course you won't have to deal with any of this on the 2.5-10 or a SWFA fixed!
Hope this helps,
-J