Rework result processing

[nck-mlcnv]

This PR mainy aims to renew the result processing as it was quite messy. Changes:

• Properties class isn't used anymore, which was quite unwieldy
• most metrics don't write RDF models anymore and only calculate single number values
• added usage of static RDF objects
• penalized metrics are now their own metrics, penalty needs to be given in the configuration now
• "new" metric "Aggregated Executions Statistics"
• the Stresstest class is now managing the result processing itself
• changed repository structure