CP algorithms: Muon trigger SF calculation improvements and fixes

Standalone muon trigger SF calculation in CP algorithms improvements and fixes:

• add ability to not save SF
• fix MC efficiency output (SF was actually written twice with different names)
• save also data efficiency
• make HLT prefix configurable -> @tstreble, why do we actually want to disable it besides the fact global trigger tool does not use it?
• make output prefixes configurable - useful if multiple WPs are covered (alternatively we could somehow add ID to the name, but current implementation allows more flexibility)
• support all years for uniform branches (year boundaries are hardcoded but probably it's fine, still I kept the default behaviour in for now)

Keeping as draft for now to get feedback but to allow more testing.

/cc @tstreble @jedebevc @ravinab