Provide API to access profiling data volume per label.

#### Is your feature request related to a problem? Please describe.

Currently is really difficult to understand what series have which profiling volume.

#### Describe the solution you'd like

Would like to provide an API, that can based on TSDB index information, determine how much profiling data has been stored.

A potential returned unit could be for example the profiling sample count. (Providing bytes, there might be tricky because of deduplication of symbols), maybe ingested uncompressed bytes would be best as that is how we charge in Grafana Cloud.

Start by locating the API entry points and TSDB index code responsible for profiling data. Clarify whether the result should report sample count, bytes, or ingested uncompressed bytes, including how labels and symbol deduplication affect it. Done means a documented API can determine profiling volume per label with tests covering the chosen metric.

Written by the indexing model from the issue text.