It's an interesting synergy.
These catalogs provide data governance, lineage, alerting, ..., and they provide it end to end, starting from extract and ending with dashboards (or even more in the future).
They usually provide agents we can trigger in our pipeline jobs - one for dbt, one for GoodData, one for Meltano, ...
We can push relevant static and runtime artifacts like dependency graphs, runtime logs, etc.
They simplify the mapping of these artifacts to users / user groups.
Users can then view status of their objects in a dependency graph and track down to find root cause of issues.
Users can also be alerted e.g. through Slack.