In this paper we compare experimental observations and theory of radiation emission from a microbunched beam with microbunching wavefront tilt with respect to the direction of motion. The theory refers to the work Tanaka et al. (2004) , which predicts, in this case, exponential suppression of coherent radiation along the kicked direction. The observations refer to a recent experiment performed at the LCLS (Nuhn et al., 2015; Lutman etal., 2016), where a microbunched beam was kicked by a bend and sent to a radiator undulator. The experiment resulted in the emission of strong coherent radiation that had its maximum along the kicked direction of motion, when the undulator parameter was detuned to a value larger than the nominal one. We first analyze the theory in detail, and we confirm the correctness of its derivation according to the conventional theory of radiation emission from charged particles. Subsequently, we look for possible peculiarities in the experiment, which may not be modeled by the theory. We show that only spurious effects are not accounted for. We conclude that the experiment defies explanation in terms of the conventional theory of radiation emission.