storage,tsdb: incorrect graphs for storage byte volumes

I believe the DB Console Storage dashboard is presenting storage metrics incorrectly. It appears to downsample using averaging (maybe a consequence of these being modeled as gauges instead of counters within the Cockroach metric system).

From the Storage dashboard:

Using sum downsampling:

Related to #102729, #75523, #99922.

Jira issue: CRDB-29274

Begin at the DB Console Storage dashboard and inspect the storage byte-volume metrics and their downsampling behavior, using the screenshots and related issues #102729, #75523, and #99922 as context. Done means the dashboard presents these volumes with correct aggregation rather than the reported incorrect graphs.

Written by the indexing model from the issue text.