String Deduplication processing

Some folks use String Deduplication and would like to observe the costs of this operation. We need to precess lines like
```
[3058ms] Concurrent String Deduplication 85/4608.0B (new), 11/352.0B (deduped), avg 28.2%, 0.065ms of 0.066ms
```

The issue names no files, tests, or entry points. Start by locating where log lines are parsed or processed, then determine the expected representation of the String Deduplication counts and timing; done means the supplied sample line is handled and its operation costs can be observed.

Written by the indexing model from the issue text.