Add a "Distinct" implementation that leverages index for low cardinality columns

For queries of the type: `select distinct() from `, we can create an implementation that enhances the query performance

No files, tests, or entry points are named. Start by tracing Pinot's handling of SELECT DISTINCT queries and how indexes are used for low-cardinality columns. Define the execution approach and validation criteria, then add tests or benchmarks showing that the stated query shape uses the index and improves performance.

Written by the indexing model from the issue text.