As-the-crow-flies (ATCF) navigation is an alternative to turn-by-turn navigation for cyclists utilizing the least-angle strategy by providing the beeline to the destination. However, past research found weaknesses (e.g., running into dead ends) affecting the user experience. In this paper, we investigate how the street network attributes to the experience of the navigation method. Using a feature importance analysis and comparison of different city types across 1633 cities, we analyze how ATCF navigation fits different environments. The perfect ACTF-city has long streets, many options to turn at decision points, few dead ends, and a grid-like structure. Cities well suited are primarily found in East Asia and North America. Furthermore, we find that previous ATCF studies were most likely conducted in Western Europe, which features the least suited street networks for the navigation method. We present design implications for future ATCF implementations and argue for diverse study locations in future research.