"You've almost answered your own question.
If the top of the lobe is reduced then the base circle must be reduced too, in order to get the same relative lobe height.
Since the donor cam wasn't worn equally on each lobe and the grinder probably removes a minimum amount of material the base heights will vary lobe to lobe."
I don't think I've explained what I'm seeing very well. The two thousands variance I was seeing is not a variance from one lobe's base circle to another lobe's base circle but rather the variance I'm seeing is on the same lobe. When the valve is fully closed, the clearance on some of valves changes as you rotate the cam. I realize that the clearance is going to change right after the valve closes and right before the valve opens but once the base circle is under the rocker arm I would think the clearance should be consistent until cam lobe comes back around to start opening the valve again.
Here's another way to look at it, if I adjust the #4 cylinder intake and exhaust valves when this cylinder is at TDC, then check the clearance on each of these valves when that valves cam lobe is 180 degrees from the rocker arm, the clearance is different, by up to about .002" and only on SOME of the valves.
When the lobe is 180 degrees from the rocker, I would think the cam under the rocker arm is at base circle.
Since (by the book) not all valves are adjusted when their cylinder is at TDC, only 1 and 4 are, it seems like the maximum valve clearance for each intake, or each exhaust valve, will not be consistent from cylinder to cylinder.
It seesm like I should not follow the book and adjust each cylinder when that cylinder is at TDC, or perhaps adjust each valve when the associated cam lobe is 180 degrees from the rocker.