Adding to that, a few things that can vary between identical airframes, that are worth checking:
- Engine power: engines vary between lumps. A cheap way to check this is tie it to a tree with a load cell in the middle, and run the engines at various powers to see what differences you get.
- Propeller pitch.
- Range of available elevator movement: this can vary by as much as a couple of degrees between "identical" airframes, so if you have a pitch authority limited stall, this could be very significant. Very easily checked.
Also, "identical" airframes, even out of the box, are not identical in dimensions, so I'd actually go and re-measure longitudinal dihedral.
There used to be an "interesting" problem with a certain Irish-built military training aeroplane, where about 1 airframe in 10 would, stalled in the landing configuration, roll itself inverted. This was eventually solved by an enterprising SEngO swapping rear fuselages and tailplanes around until the problem went away. The basic problem was tolerances in design adding up to a cumulative significant difference in shape.
G