[EPIC] Performance focus for 0.2.0 Release #717

What is the problem the feature request solves?

I am creating this epic to track performance issues that the community is working on for the 0.2.0 release.

• Investigate options for improving performance when reading decimals from Parquet #679
• Investigate performance of decimal math / aggregation #670
• Evaluate performance improvement from fusing aggregate and filter to use avoid copying the batch (see Fuse grouped aggregate and filter operators for improved performance datafusion#5944 (comment))
• Evaluate use of selection vectors in scan-filter-join operations #745
• Improve performance of broadcast hash join #808

Describe the potential solution

No response

Additional context

No response