Memory models are subtle. Temporal reasoning in context of h/w, e.g. multi-core coherence, and language runtime MM is non-trivial. So there is a baseline level

Memory models are subtle. Temporal reasoning in context of h/w, e.g. multi-core coherence, and language runtime MM is non-trivial. So there is a baseline level of difficulty baked into the domain. Education certainly is necessary, but here can only inform of what needs to be considered, known pitfalls, patterns of concurrency, etc. As to OP, well it better be viable, because we certainly need to deal with it. So better tooling and conventions encapsulated in expert developed libraries. The education level required will naturally fall into the categories for those who will develop the tools/libraries, and those that will use them.