Yet the question maybe related to what happens in a prestressed column: when you compress it, the compressive stress is increased, and some of the tensile stress in the prestressing steel, lost. So it is clear that once the transfer has taken place (and assuming grouted tendon or adhered) the whole section, steels included needs be taken for the computation of the center of gravity of the shape, to serve in the elastic calculation of the member.
Before prestress and its grouting, only the passive rebar and concrete section is usually taken unto account for finding center of gravity of the section, the acting prestress being if wanted substituted bu equivalent forces to act unto such pre-existing shape.
Postensioning in a non grouted duct is akin to this, but the integrity of the section can't be assumed. SO for this case again, only passive rebar plus concrete would give the center of gravity, the prestress being substituted by its equivalent forces...
what brings us to the conclusion of that to consider the prestressing steel part of the section to the effect of the determination of mechanical properties to be used in elastic analyisis of the section and member, it is extant bonding what determines its inclusion.
The non bonded cases, to be treated as forces acting on what bonded.