RE: Trying to understand incidence
Getting back to models, if you look at some of the "classic" Pattern designs, like the Joe Bridi Super Kaos, the wing had some positive angle when referenced to the fuselage datum, and the engine had some down angle. The stabilizer was at zero. These angles gave the model the desired effect of needing virtually no stabilizer trim in upright level flight, and a small amount of down elevator when inverted, at the speeds flown in the then-current AMA Pattern competition.
Naturally, the design was intended to fit Joe Bridi's style of flying the AMA Pattern at the time, so what fit his style might not have fit other flyers. Even so, the airplane had become known over time as an excellent flying airplane, and variants are still being produced today, over thirty years later!
On the Super Kaos, the fuselage top would be leveled and used as the angular reference, and the wing's, stabilizer's, and engine's angles would all be measured in relation to that. Level the fuselage and then use an incidence meter (angle meter?) to set the wing, stab, and engine.
As has also been stated, the fuselage reference lines (FRL---I've learned to call them datum lines) are arbitrarily-set by the aircraft's designer to have a convenient reference to make all other relative measurements. On the Van's Aircraft RV-9 I'm building, the longitudinal FRL is the top of the top fuselage longeron, and the vertical FRS is defined just forward of the propeller spinner. To make it easier on the home aircraft builder, the wing's and stabilizer's angles are measured in inches from the datum line at the leading edge and trailing edge of the wing and stab.