The Immutable Priority Queue has poor performance in specific situations

[Lampese]

The immut/priority_queue in the current core uses the pairing-heap algorithm, which only folds the entire heap when querying the results. Other operations are O(1). Although its constant is very small and the average complexity can be O(log n). But if it is in the case of immutable, it is likely to cause too much time to be spent on a certain version of the operation, resulting in an invalid average complexity. We should use algorithms that can maintain stable complexity even in immutable situations.