Scale scoring within each result set

[karafecho]

This issue is to suggest that we scale the scoring within each result set instead of removing all scores, which apparently is a decision under consideration by the SA1 team. FWIW, in my experience, users aren't concerned about the absolute score once you explain that it's actually a relative score. See #281.