Drones hold promise to assist in civilian tasks. To realize this application, future drones must operate within large cities, covering large distances while navigating within cluttered urban landscapes. The increased efficiency of winged drones over rotary ones is advantageous. However, the applications of winged drones are so far limited to open spaces, as significant challenges arise when introducing them to urban environments. For winged drones to be viable in such environments, they must be able to navigate with high accuracy and agility. Seeking inspiration from nature to achieve this goal, we look to birds of prey. These birds are renowned for their ability to travel long distances and for their accurate, agile hunting abilities. Birds of prey utilize their sharp vision for accurate navigation and morph their wings to adjust to momentary needs and achieve their high agility. As a result, researchers have explored morphing wing drone designs. However, to unlock their potential in agility and accuracy, advanced control techniques are required for these hard to model and control drones. To improve accuracy, we propose the integration of vision, as seen on birds of prey. By combining satellite data with vision, we bring together long-range navigation and close-range precision. In our experiments, we outperform current approaches and increase accuracy over only satellite data flights by over an order of magnitude. To achieve agile flight for winged drones, the development of computationally and sample-efficient control systems is crucial. Computational efficiency is vital to reduce reaction time and payload, which directly affects flight range. Sample efficiency is crucial for adapting to the complex, hard to model dynamics of these drones. To obtain computationally efficient controllers that allow data-efficient adaptation to address modeling mismatches, we investigate a novel advanced control approach that leverages model knowledge. We further build on insights from birds by employing avian flight strategies to achieve flight control for agile morphing drone flight. We demonstrate this on an agile perching maneuver, where we demonstrate the importance of morphing wings in achieving such a maneuver. Traditional aircraft design principles are however of limited applicability to morphing wing drones whose performance is directly influenced by control strategies. Therefore, we propose a closely integrated design and control approach to optimize performance and demonstrate enhanced energy efficiency and reduced required flight time.