That fourth point is a great one.

One of the most frustrating things in software engineering is eyeballing a decision (i.e. naming a table with a plural) and just gut-level knowing it's the wrong decision, but not quite being able to remember specific pitfalls that instilled that feeling earlier in your career.

That fourth point is the kind of thing that just cuts right through the conversation - "doing it this way does not scale and will fail, here's an example of how". Such frission in finding those and getting consensus around them, because you can feel the future pain you've avoided evaporating.