Push additional parquet filtering into the parquet scan [EPIC]

[alamb]

Is your feature request related to a problem or challenge? Please describe what you are trying to do.
The more filtering that can be pushed to the parquet reading, the faster a query will run in general as less work is needed to decode and process data that will eventually be filtered from the plan

There are several ongoing workstreams that will eventually lead to pushing down substantial additional filtering into the parquet scan that should substantially increase performance for datafusion. I wanted to capture them here to provide more visibility

cc @Ted-Jiang @tustvold @thinkharderdev

Describe the solution you'd like
Here are some of the tasks I have collected. There are likely more -- please add them (either directly or via comments)

• Switch to object_store crate (#2489) #2677
• Parquet Scan Filter arrow-rs#1191
• Changes to ParquetRecordBatchStream to support row filtering in DataFusion arrow-rs#2270
• Implement parquet page-level skipping with column index, using min/max stats #847
• Write a blog post on the topic

[alamb]

Duplicate of #3462