Implications and limitations of Townsend’s attached-eddy hypothesis are considered for finite Reynolds number wall-bounded flows. Townsend’s hypothesis forms the basis of an established model of the logarithmic layer, which is characterized by a hierarchy of geometrically self-similar eddying motions scaling with their distance to the wall. Current anomalies with the theory and empirical observations for the logarithmic growth in the u-variance are considered, including the related mismatch in the k⁻¹ scaling pre-factor of the streamwise velocity spectra for streamwise and spanwise wavenumbers (k). We also briefly revisit the mathematical basis of the attached eddy hypothesis to reveal how fundamental quantities such as the von Karman constant are related to the geometry and mean-spacing of the statistically representative attached eddies. The implications of this for different pressure-gradient wall-bounded flows will be discussed in light of recently conducted high-Re APG experiments at Melbourne.