query caching leads to absurdly slow queries [LUCENE-9002]

**Description**
We have dozens of ES clusters(based on Lucene) for metric scenarios. Most of the queries are like this: _host_ip:10.10.10.10 AND timestamp:[2019-10-01 00:00:00 TO 2019-10-05 23:59:59]_. And we frequently encounter some absurdly slow queries.

**Solution**
For a long time range query(e.g. 5 days), each range query will consume tens of megabytes of memory and spend hundreds of milliseconds to cache, but the benefits are not obvious. And those large cache entries will cause frequent cache eviction. So it's better to skip the caching action directly when large range query appears with a selective lead iterator.

---
Migrated from [LUCENE-9002](https://issues.apache.org/jira/browse/LUCENE-9002) by Guoqiang Jiang, updated Aug 11 2021
Linked issues:
- #9075
- #9260

Start by reading the migrated LUCENE-9002 discussion and the linked issues #9075 and #9260, then trace how long range queries enter query caching. Confirm the memory and cache-eviction behavior described for selective lead iterators, and define completion as avoiding ineffective caching without regressing query results or useful cache hits.

Written by the indexing model from the issue text.