CTAS: Erroneous "Duplicate values found in schema"

**Describe the bug**

The assignment of internal column names is not consistent across different queries. The problem might be in the auto-increment when creating a new internal column name.

**To Reproduce**
Steps to reproduce the behavior, include:
1. The version of KSQL.
Master
2. Sample source data.
No need for source data
3. Any SQL statements you ran

First create a table with a value column without any alias. Then, try to create a new table that uses this column again without an alias.

```
describe metrics_cube;

Name : METRICS_CUBE
Field | Type
---------------------------------------------------
ROWTIME | BIGINT (system)
METRICS_STREAM.ROWKEY | VARCHAR(STRING)
METRICS_NAME | VARCHAR(STRING)
METRICS_VALUE | DOUBLE
KSQL_COL_2 | ARRAY
---------------------------------------------------
For runtime statistics and query details run: DESCRIBE EXTENDED ;
```

```
CREATE TABLE metrics_table AS
SELECT metrics_name, KSQL_COL_2, avg(metrics_value) from metrics_cube WINDOW TUMBLING (size 60 seconds)
GROUP BY metrics_name, KSQL_COL_2;
```

Gives the error
`Duplicate values found in schema: `KSQL_COL_2` DOUBLE`

**Expected behavior**
Assign a new internal column name `KSQL_COL_3` to the non-aliased column

**Actual behaviour**
A clear and concise description of what actually happens, including:
1. CLI output
2. Error messages
`Duplicate values found in schema: `KSQL_COL_2` DOUBLE`
3. KSQL logs

**Additional context**
Add any other context about the problem here.

No files or tests are named. Start by reproducing the supplied CTAS statements against the described schema and trace the internal column-name assignment for the non-aliased KSQL_COL_2 reference. Done means the second table is created without a duplicate-schema error and the column receives the expected KSQL_COL_3 name.

Written by the indexing model from the issue text.