Combining both articles and sprinkling in some of my own experience, I would state the following concepts/rules to consider:

1. Pragmatism over Novelty: Both articles advocate for prioritizing practicality and reliability over the allure of new and trendy technologies. They argue that while new technologies may seem exciting, they often come with risks and complexities that can outweigh their benefits. At Toreon, when teaching threat modeling courses, pragmatism is always emphasized—use the techniques that work well, discard everything else (this assumes you understand your own threat modeling process well).
2. Focus on Stability and Reliability: McKinley and Allspaw stress the importance of choosing technologies and solutions that are proven to be stable, reliable, and well-understood. They argue that stability and reliability are critical for the smooth functioning of systems and minimizing disruptions. This also simplifies the creation of a threat model considerably; new types of attacks still need to be analyzed, but many security problems have already been solved by the community and do not require extensive analysis anymore.
3. Consideration of Multiple Perspectives: Allspaw’s article emphasizes the importance of considering diverse viewpoints and insights when analyzing technical problems and designing solutions. Similarly, McKinley’s article indirectly supports this idea by advocating for a pragmatic approach that considers various factors, including stability, support, and business needs. It is my experience that the group of people needed to create a good threat model should include at least a person who understands the business (could be a product owner or a business analyst), an application architect, and a lead developer. There are also times when you will need stakeholders from the infrastructure team, legal team, and more.
4. Holistic Understanding: Both authors suggest that a holistic understanding of technical challenges is essential for effective problem-solving. McKinley emphasizes the need to focus on solving actual problems rather than chasing the latest trends, while Allspaw highlights the value of integrating insights from different disciplines to gain a comprehensive understanding of complex systems. This element cannot be underestimated; there are many different security activities that can be performed when creating a secure product, all aimed at the same goal: creating a more resilient and secure product. I believe a good threat modeler has a deep understanding of all these activities. (Hint: OWASP SAMM) is an excellent resource for discovering these different security activities).