Abstract. In a recent paper, which presents CCD photometry for fifteen dwarf and intermediate early-type galaxies in the NGC 5044 Group, it has been claimed that "a few relatively bright galaxies with "convex" profiles destroy the known relation between total magnitude and the "shape" parameter... thus ruling out the use of this relation as a distance indicator for individual galaxies". In the same paper, further reasons were cited supposedly "limiting also its use as a distance indicator for groups of galaxies". We demonstrate that none of the three relatively bright galaxies cited as possessing "convex" profiles actually has a convex profile, and that one of these objects should be excluded because it is a late-type galaxy. Of the two remaining objects, one has an anomalous profile shape whilst the other is brighter than one might expect from its colour alone. However, we show that all of the other issues raised have already been accounted for by Young & Currie (1994, 1995. The main implications of the new observations are: (1) that the case of one galaxy with an anomalous profile shape, N42, highlights the need for some a priori criteria to be defined in order to establish objectively which objects are not suitable for distance determinations; and (2) on the basis of another unusual galaxy, N50, colour has now been shown to be a poorer discriminant between objects of the same profile shape and scalelength (but of different central surface brightness) than previously thought. How significant this latter problem is depends on how common N50-like objects are. This consideration reinforces the case for always using the more general scalelength-shape relationship of Young & Currie (1995) in preference to the luminosity-shape one of Young & Currie (1994). Reassuringly, through a re-analysis of the same CCD photometry, we find that NGC 5044 Group galaxies observe a tight scalelength-shape relationship. This finding supports the view that the scalelength-shape relationship is a viable distance indicator.