Ask HN: Role of an SRE in a Reliable Service

What's the role of an SRE on a team that owns a fairly reliable service that has identified scaling concerns, alerting/monitoring is well setup, and is managed by a well run team? Obviously, there could still be some unknown issues, or misses, but perhaps not enough work for a full-time SRE in that team? Does it make sense for SREs to be switching their focus to writing business logic for new features/APIs until there's more "SRE"-type work available?