As extended reality (XR) technologies become increasingly integrated into various industries, understanding their potential and limitations is crucial for professionals to understand the best use cases of this technology. One of the methods that can be used to assess the usability of these devices is a heuristic evaluation, a subjective interface analysis method based on a set of principles or rules of thumb. Derby (2023) validated a new heuristic checklist (Derby Dozen) specifically for AR and MR devices and applications, with 12 heuristics and 109 checklist items. This study will present this heuristic checklist and show a comparison of the Apple Vision Pro (AVP) and the Meta Quest 3 (MQ3) for a series of applications and tasks. The evaluation compared eight of the twelve heuristics of each device, after completing a series of tasks including setting up the headset, using two types of interactive apps, typing, and changing the brightness level. Results from the heuristic evaluation suggest that both the AVP and MQ3 demonstrate generally positive usability performance, with the MQ3 receiving superior ratings across most heuristics. Both headsets exhibited numerous violations of the heuristics of comfort and intuitiveness of virtual elements. The AVP was noted for its high-quality real-world and enjoyable eye-tracking-based interaction method, although it faced limitations including feedback delays and physical discomfort. The MQ3 offered flexibility in its interaction methods and stronger system feedback, but also presented challenges such as physical discomfort and hand-tracking inconsistency. Design implications, study limitations, and directions for future work are discussed.