Quantum coherence is proclaimed to promote efficient energy collection by light-harvesting complexes and prototype organic photovoltaics. However, supporting spectroscopic studies are hindered by the problem of distinguishing between the excited state and ground state origin of coherent spectral transients. Coherence amplitude maps, which systematically represent quantum beats observable in two-dimensional (2D) spectroscopy, are currently the prevalent tool for making this distinction. In this article, we present coherence amplitude maps of a molecular dimer, which have become significantly distorted as a result of the finite laser bandwidth used to record the 2D spectra. We argue that under standard spectroscopic conditions similar distortions are to be expected for compounds absorbing over a spectral range similar to, or exceeding that of the dimer. These include virtually all photovoltaic polymers and certain photosynthetic complexes. With the distortion of coherence amplitude maps, alternative ways to identify quantum coherence are called for. Here, we use numerical simulations which reproduce the essential photophysics of the dimer to unambiguously determine the excited state origin of prominent quantum beats observed in the 2D spectral measurements. This approach is proposed as a dependable method for coherence identification.