Rework scan result handling for oscap

[comps]

Currently, we parse oscap --verbose=INFO --progress output and report individual rules in realtime.

This gives us info on how oscap treats OVAL checks for that rule, but without more context (which files failed the check, etc.), this is useless.

Instead, let oscap generate ARF results as a XML and parse it after the scan - this allows us to

1. get more context for file path related checks and maybe include it in the note (which checks failed)
2. parse out remediation shell script output and submit it for each rule as a separate log

Especially the second point is critical for investigating seemingly random fails as those are often caused by the remediation script running into unforseen environment setups.