Submissions statistics and following problems

```
Report:
Number of files: 3250000
Average file size: 4939.22 bytes

Average per submission:
checkouts: 0.48
builds: 0.51
tests: 2.04
issues: 0.00
incidents: 0.00

Total counts:
checkouts: 1559720
builds: 1669072
tests: 6619087
issues: 4308
incidents: 945

Total duplicate IDs: 324327

Versions:
4.0
4.2
5.0
5.1
5.3

Submitters:
broonie: version 4.0
maestro: version 5.3
linaro: version 5.3
redhat: version 4.2
riscv: version 5.1
microsoft: version 5.0
ti: version 5.1
```
As we can see, we have certain problems related to multiple different schema versions submitted. We need to figure out best way to handle that without "automagic" kcidb conversion.

Start with the issue's submission statistics and compare the listed schema versions and submitters. Investigate how differing schema versions are currently handled and define an approach that avoids automagic kcidb conversion. Done means the project has an agreed, implementable way to handle submissions from multiple schema versions.

Written by the indexing model from the issue text.