Have a question - what would be the difference between data lake ingestions on s3 files vs getting this from Glue?

I think Data Lake ingestion is going to pull the more raw S3 object definitions, where Glue adds some additional metadata on top to formally turn them into a queryable "dataset"