Just wraps a Scorer and performs top scoring using it.
Same as DocIdSetIterator.cost for bulk scorers.
Collects matching documents in a range and return an estimation of the next matching document which is on or after max.
max