Spark SQL MIT global index change partition paths will not work with custom merger

We use avro record merger so we can use expression payload, and we overwrite the merger impls. There might be some changes to how we do merger impls so I don't want to address this issue yet. 

If overwrite or event time is used it is fine because we just use the payload based merging. But if they have a custom merger, then we need to use that when reading the records