One aspect I see rarely discussed is matching technology to the skill of the team. IMO people should not choose technologies they're unwilling to become experts

One aspect I see rarely discussed is matching technology to the skill of the team. IMO people should not choose technologies they're unwilling to become experts with, or lack the workforce capable of. It's a hard fact of life that some are just more capable than others (either through intrinsic factors like "Intelligence", or extrinsic like family circumstances, quality of education received et al.). Obviously the extrinsic ones are in our locus of control, so an engineering team that presently is incapable, can grow to fill that need. I don't think most companies w/ a high requirement for quality (say something that affects others' health for example) should use something like AWS unless they have a workforce dedicated to understanding distributed systems (academically) and become an expert in each AWS technology they choose. I have observed so many engineers treating these tools w/ their wishful assumptions of how they ought to work versus how they actually work... There's probably a similar conversation to be had about databases' guarantees, consensus algorithms, amongst others...