Automatic fatigue crack detection using commercial sensing technologies is difficult due to the highly localized nature of crack monitoring sensors and the randomness of crack initiation and propagation. The authors have previously proposed and demonstrated a novel sensing skin capable of fatigue crack detection, localization, and quantification. The technology is based on soft elastomeric capacitors (SECs) that constitute thin-film flexible strain sensors transducing strain into a measurable change in capacitance. Deployed in an array configuration, the SECs mimic biological skin, where local damage can be diagnosed over large surfaces. Recently, the authors have proposed a significantly improved version of the SEC, whereby the top surface of the sensor is corrugated in diverse non-auxetic and auxetic patterns. Laboratory investigations of non-auxetic patterns have shown that the use of corrugation can increase the sensor's gauge factor, linearity, and signal stability when compared to untextured sensors, while numerical analyses of auxetic patterns have shown their superiority over non-auxetic corrugations. In this paper, we experimentally study the use of corrugated SECs, in particular with grid, diagrid, reinforced diagrid, and re-entrant hexagonal honeycomb-type (auxetic) patterns as a significant improvement to the untextured SEC in monitoring fatigue cracks in steel specimens. Results show that the use of corrugation significantly improves sensing performance, with both the reinforced diagrid and auxetic patterns yielding best results in terms of signal linearity, sensitivity, and resolution, with the reinforced diagrid having the added advantage of a symmetric pattern that could facilitate field deployments.