At the turn of the decade we arguably move from the era of precision cosmology to the era of galaxy formation. One approach to this problem will be via the construction of comprehensive galaxy samples. In this review I take the opportunity to highlight a number of challenges which must be overcome before we can use such data to construct a robust empirical blueprint of galaxy evolution. The issues briefly highlighted here are: the Hubble tuning fork versus galaxy components, the hierarchy of structure, the accuracy of structural decompositions, galaxy photometry, incompleteness, cosmic variance, photometric versus spectroscopic redshifts, wavelength bias, dust attenuation, and the disconnect with theory. These concerns essentially form one of the key motivations of the GAMA survey which, as one of its goals, will establish a complete comprehensive kpc-resolution 3D multiwavelength (UV-Opt-IR-Radio) database of 250 k galaxy systems to z<0.5.