We have previously shown that both, polymer conformation and dynamics are crucial for the exciton transport in conjugated polymers. Thereby we found that the usual Förster-type hopping transfer model – even if one applies the line-dipole approximation – falls short in one crucial aspect: the nature of the sites the excitons are transferred between is still unclear. We found that the simple model of spectroscopic units defined as segments of the polymer chains separated by structural defects breaking the π-conjugation is only justified for chemical defects like hydrogenated double bonds, or extreme gauche (90°) torsions between the monomers. Both defects are far too rare in a well-prepared conjugated polymer to explain the mean spectroscopic-unit length of typically 6–7 monomers. Meanwhile, also the concept of dynamical formation of the spectroscopic units, we had previously suggested, has also failed. Thus the question of a paradigma change concerning the exciton transport in conjugated polymers appears on the agenda.