More fluent pipeline

[Hendrik-code]

• Allow matching to yield all computed metrics to the instance evaluator
• Allow matching to give matches based on multiple thresholds (so instance evaluator can immediately calculate AUTC values)
• The matcher should only yield the changed labels, and a matching score so that afterwards, we can do the AUTC thing.
• Basically always let the instance evaluator calculate all matches based on threshold 0 (maximum number of matches) and then have a function that yields this for a threshold x and over all thresholds as AUTC.
• Save matched instances already as cutouts, should improve speed by a lot.