The additional expense in OP had nothing to do with scale; it arises from: redundancy, analytics, off-the-shelf integrations.
It is a common misunderstanding amongst software engineers that infrastructure serves "performance". Most of the complexity comes from redundancy and analytics (, realtime especially).
Right, but how much do these nice-to-haves add to the revenue? Is it truly worth it to bake them all in from the very beginning? E.g. five minutes of downtime in morning hours will affect only a handful of users, and if so, why bother with blue-green deployments other than out of professional interest?
It is a common misunderstanding amongst software engineers that infrastructure serves "performance". Most of the complexity comes from redundancy and analytics (, realtime especially).