I may not be picturing your set-up perfectly, but I'll dive in nonetheless:
The big picture is that the backing disk of a magnet rotor is part of the flux path. By putting magnets on the other side, you are disturbing the flux path of the magnets on the other side, and they will also disturb the flux path of the magnets you just added. I'm sure you know this (or you wouldn't ask) but I think you're looking for a quick way to set up a few tests, just to get the coil winding right and move on...
Perhaps you could make a rough estimate that, if you did everything else perfectly, then the contribution of the backing plate to each group of magnets is 1/2 of what it should be. That probably isn't true either, because the flux density in the rotor disk cannot increase beyond the saturation limit, and the first magnets probably already did that. Unless the backing disk is very thick. Perhaps it would have to be 2x as thick as the magnets themselves.
Even if there is no measurable leakage on the back-side (can a paper-clip stick?) the flux density could be high enough that your experiment could be invalid. Depends on what you want from the results, though - if it's just to make some quick comparisons, then this is still a go. If you intend to get detailed and mathematical then it's probably the wrong starting point.