It doesn't really matter if one way is stable and the other way is unstable, so long as the stab is roughly 0deg at the tipping point then it's accuracy is good plus or minus a few mm.

Most will argue that determining exactly where the CG is ranks far lower than actually having it in the best location determined by flight testing. It's a different story if you're swapping different weight batteries and need to reproduce an exact location for consistency. For that, I'd get it close, flight test and mark the individual battery's location.

Once you know the rough CG location, then using a calculator to figure out any changes takes all the guess work and heavy(?) lifting out of it.