Update chunked parquet reader benchmarks

[sdrp713]

Description

This PR addresses issue #15057 by contributing the following:

• Adds a new benchmark that provides nvbench axes for both chunk_read_limit and pass_read_limit
• Renames byte_limit to chunk_read_limit in BM_parquet_read_chunks
• Adds an nvbench axis for data_size to allow the benchmarks to operate on tables larger than 536 MB.

Checklist

• [x] I am familiar with the Contributing Guidelines.
• [x] New or existing tests cover these changes.
• [x] The documentation is up to date with these changes.