That's interesting. My kneejerk reaction to your comment was "no, you don't have a strategy for landing a man on the moon, you have a mission architecture!". But on further consideration there really seems to be a lot of overlap between system architectures and strategies (in the Rumelt sense). Diagnosis, kernel, guiding policies. Seems to fit both domains.
this is something i'd really like to read about, and how they design for all the contingencies and things that can go wrong on a practical and hardware/software level.