Many crucial activities in software development are linked to gaze and can potentially benefit from gaze-assisted developer tools. However, despite the maturity of eye trackers and the potential for such tools, we see very few studies of practitioners. Here, we present a systematic mapping study to examine recent developments in the field with a focus on the experimental setup of eye-tracking studies in software engineering research. We identify two gaps regarding studies of practitioners in realistic settings and three challenges in existing experimental setups. We present six recommendations for how to steer the research community toward gaze-assisted developer tools that can benefit practitioners.