Practical degreee of accuracy thicknessing panels

[Deems Davis]

After having my shop packed for 2 1/2 years due to move and new home construction, I'm finally back to dabbling in guitars. So I've been back thru the "books" and reacquainting myself with the theories and processes. I've recently been down the rabbit hole of the Excel spreadsheets, and found that for the life of me I couldn't EXACTLY match the example results. FINALLY I found the threads that explain that others with similar problems and that the final results differences were a computing phenomena and NOT material as +/= .03-.05 mm wasn't PRACTICAL as to achievable accuracy.
So I've accepted my results and am thicknessing the panels.
The degree of accuracy implied in the formulas e.g. .000000x... leads folks like me to assume that there is a NECESSARY level of precision in skills in order to obtaining valid final results.
I now have it thru my thick head that the theoretical and the practical are not necessarily identical (at least for me and my lifetime).
So as I'm thicknessing my panels (particularly the softer spruce tops). I'm finding some variations ( +/- .0x in in measurements as I sample 20 odds points in the panel.) I should add that I'm using a drum sander.
What degree of accuracy (+/-} are you more experienced folks finding is a PRACTICAL acceptable deviation? Say you are targeting a top to be 2.5 mm thick how many sample points do you check on the panel? and how much of a variation can you accept as practically achieving the target? Or am I the only one thats not able to turn out perfectly uniform results?

Deems

[Mitch Lees]

I find it quite difficult to level my Jet drum sander to better than .2 to .3 mm. Usually one end or the other is just slightly high (more often the motor end). Reversing the feed of the boards each pass helps, but usually I find .1 or .2 variation. I note the high side and take just a little more off during the edge thinning process.

Mitch

[Trevor Gore]

The +/- 0.05 is practical for me, using planes. On the occasions that I've used Gerard's wide belt sander, the variation is a lot more than that, though, as it also is on the (few) typical-style cantilevered drum sanders I've had experience with.

In order to get good tolerance with a plane, you need a very flat and smooth planing board, sat on a flat bench. If you then plane with a longish (#5 1/2 or 6) flat plane (it's easier to achieve flatness with the wide planes) the +/- 0.05mm is not hard to achieve.