RE: Trying to understand incidence
Ben,
I agree with your statement about having a line outside of the aircraft, especially with a host of other people involved in the process. That would make things easier in a team setting.
In the case of the RV-9, Van's has one person who did the actual design. Since they then produce kits, the references used make it very easy for the builder to get things aligned. All that's needed for alignment are a good level, an accurate ruler, and a plumb bob. The components are put on a stand/jig and levelled and plumbed as needed...shim as necessary and set the clamps. For example, when I set up the wings for drilling the rivet holes, I set it up on a stand that had the chordline vertical with the wing spar horizontal. A plumb bob was hung from the main spar and the distance from the plumb line to the rear spar was measured. As long as the root and tip had less than 3/32" of twist, the wing was considered good (over a 12' span in the wing panel). Of course the object of the homebuilder is to get things as close to perfect as possible.
With respect to models, something similar is happening. The designer has to make it easy for the modeler to true-up the airframe. Having a straight-line reference on the model makes it easy. A fuselage longeron, top deck, crutch, and so on as the reference make it easy to set the model and get it levelled in the building cradle. An "incidence meter" can then be used to get the wing and stabilizer (engine thrustline, too) to the angles detailed in the instrucions/plans. If the datum line was a foot below the model, it would be much more difficult for the modeler to set up the alignment.
Ultimately, the actual location of the reference is irrelevant, as long as it makes it easy for the builder/assembler to get the job done with the least possibility of error. As a long-time modeler, I know I make enough of them on my own!