dfcae7b created the issue

repro: upload some gzipped plaintext with content-encoding "gzip" and content-type "text/plain"
cyberduck download-to: file is not decompressed
cyberduck copy url and download from browser: file is decompresed properly

I think this was introduced to fix something else (#8263)

In the linked ticket I do not think the root cause was understood correctly - gzip content-encoding and s3 is fine, however Cloudtrail logs used concatenated gzip and that is not well supported for content-encoding (every current major browser will behave the same and end the response after the first gzip file. Opera is the exception - see this relevant bug report for Chrome marked wontfix https://bugs.chromium.org/p/chromium/issues/detail?id=20884)

expected behavior: I would prefer files with content-encoding are decoded accordingly, and wouldn't mind an additional option to download files unmodified for special use cases or as a work around for Cloudtrail's buggy behavior.