Targeted nanomedicines are commonly evaluated through measurements that operate at different biological scales but are often interpreted as though they describe the same delivery event. Blood pharmacokinetics, organ biodistribution, local tissue accumulation, cellular association, internalization, intracellular trafficking, and payload availability answer different questions. This Perspective argues that the central translational problem is therefore not simply how much nanomaterial reaches an organ, but how exposure is partitioned across successive biological compartments and carrier states. Evidence from tumour nanomedicine, lipid nanoparticles, extracellular-vesicle systems, biomaterial transport studies, and intracellular imaging shows that major losses and redistributions can occur at vascular passage, tissue penetration, cell-type selection, uptake, and intracellular release. We propose that targeted-nanomedicine development should use multiscale exposure maps that retain these states separately and connect them only when carrier identity, biological compartment, measurement scale, and time are sufficiently aligned. Within this logic, molecular affinity becomes informative only after the carrier or its functional payload has obtained access to the relevant biological interface. The framework is not presented as a validated universal pharmacokinetic model. Rather, it is a measurement and interpretation architecture intended to expose where apparent delivery success reflects organ accumulation, where it reflects genuine cellular access, and where missing measurements prevent those states from being distinguished. Such maps could support more disciplined formulation comparison, mechanistic interpretation, and translational decision-making while preserving model-system and assay boundaries.