You are advising a customer running hourly Spark ETL jobs on Databricks and need to prototype a small simulator that chooses cluster size over time. Write a program that, given a sequence of hourly workloads, a latency SLA per hour, and per-node hourly cost, computes a scaling plan that minimizes total cost while keeping estimated job runtime within SLA. Assume runtime scales sublinearly with node count according to a provided formula, and include a configurable startup penalty when scaling up. After producing the plan, print total cost, any SLA violations, and a short explanation of why the chosen plan is cost-efficient. In the discussion portion, be prepared to explain how your approach would change for an executive audience focused on business tradeoffs rather than implementation details.