RF: Adjust argument order for evaluation routines so they can be used by `feedsRun` scripts
The pyfeeds.evaluate.eval* functions can now be called with just test file and benchmark file names. Not sure why I didn't do it this way in the first place.