Explain the concept of column-level statistics and histograms in query optimization and cost-based query planning, and how they inform the query optimizer about data distribution, selectivity estimates, and cardinality estimates for efficient query execution?