Dear Ludwig,
Regarding your first question, here the definitions of the aerodynamic (AeroTwst) and structural (StrctTwst) twist that we use in FAST/AeroDyn (and in the NREL 5-MW turbine documentation):
AeroTwst - Rotation of the chordline of the airfoil relative to the zero-twist location, such that the blade pitch angle plus AeroTwst is the angle between chordline and the rotor plane (neglecting structural deflection such as elastic twist).
StrctTwst - Rotation of the principle elastic axis for edgewise bending of the airfoil relative to the zero-twist location, such that the blade pitch angle plus StrctTwst is the angle between principle elastic axis for edgewise bending and the rotor plane (neglecting structural deflection such as elastic twist).
So, AeroTwst and StrctTwst would be identical if the principle elastic axis for edgewise bending lied along the chordline.
Regarding your second question, there are some discussions about interpolating airfoil geometry versus interpolating airfoil data in the forum topic found here: Airfoils on a 5m Diameter Turbine.
Regarding your third question, I agree that sounds strange, but I don’t have an answer.
Best regards,