Don't use BooleanScorer for small segments [LUCENE-7365]

If a BooleanQuery meets certain criteria (only contains disjunctions, is likely to match large numbers of docs) then we use a BooleanScorer to score groups of 1024 docs at a time. This allocates arrays of 1024 Bucket objects up-front. On very small segments (for example, a MemoryIndex) this is very wasteful of memory, particularly if the query is large or deeply-nested. We should avoid using a bulk scorer on these segments.

---
Migrated from [LUCENE-7365](https://issues.apache.org/jira/browse/LUCENE-7365) by Alan Woodward (@romseygeek), updated Jul 04 2016
Attachments: [LUCENE-7365.patch](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-7365/LUCENE-7365.patch) (versions: 3), [LUCENE-7365-query.patch](https://apache.github.io/lucene-jira-archive/attachments/LUCENE-7365/LUCENE-7365-query.patch)
Linked issues:
- #8418