DB-Level Changefeed: Log when a restart occurs, and include cause (table add/remove, schema change, etc.)

Having the restart cause will help us to monitor when things are going wrong with which tables out changefeed targets vs. for other reasons.

Epic: [CRDB-1421](https://cockroachlabs.atlassian.net/browse/CRDB-1421)

Jira issue: CRDB-51250

Epic CRDB-1421

Start by locating the DB-level changefeed restart handling and any existing restart logging. Trace the paths that restart a changefeed, identify how table changes and schema changes are represented, and confirm the expected cause categories and log fields with the issue discussion before implementing; done means each restart records an actionable cause.

Written by the indexing model from the issue text.