Unnecessary cartesian explosion if multi-value column is reused in expression

### Affected Version

Druid 0.16, 0.15

### Description

With this dataset (as `fs`):

```json
{"time":"2019-08-14T00:00:00.000Z","srcGroups":["x","y","z"],"dstGroups":["a","b","c","d"]}
{"time":"2019-08-14T00:00:00.000Z","srcGroups":["x","y","z"],"dstGroups":["a","c","d"]}
{"time":"2019-08-14T00:00:00.000Z","srcGroups":["x","y","z"],"dstGroups":["a","g"]}
```

Doing this query:

```sql
SELECT
CASE "dstGroups"
WHEN 'b' THEN 'b'
WHEN 'g' THEN 'g'
ELSE 'Other'
END AS "dst",
COUNT(*) AS "Count"
FROM "fs"
GROUP BY 1
ORDER BY "Count" DESC
```

Yields an unexpected cartesian explosion:


This is due to this expression in the underlying plan:

`"case_searched((\"dstGroups\" == 'b'),'b',(\"dstGroups\" == 'g'),'g','Other')"`

Which triggers a cartesian product on the same column `dstGroups` which it should not do.

Reproduce the issue with the provided fs dataset and SQL query, then inspect the generated plan containing the case_searched expression. Trace why reusing dstGroups creates a cartesian product; done means the query returns expected grouped counts without that explosion.

Written by the indexing model from the issue text.