By looking at the VPC flow integration, we are missing bucket_arn and number_of_workers in the configuration list:

Also when choosing to collect VPC flow logs from S3, s3 bucket information should also be stored into each document. These fields are collected in Filebeat but not registered in the integration.