DataFlow Data freshness metric definition

I'm struggling to get the exact definition of data freshness because it is defined differently in several places.
This is a hint from the dataflow UI:

"The number of seconds since the most recent watermark." - It's easy, like now minus the output watermark.

But extended doc says:

"The data freshness metric shows the difference in seconds between the timestamp on the data element and the time that the event is processed in your pipeline. The data element receives a timestamp when an event occurs on the element, such as a click event on a website or ingestion by Pub/Sub. The output watermark is the time that the data is processed."

And this is confusing. It doesn't seem logical that "output watermark is the time that the data is processed." basically if output watermark is equal to processing time it means the queue is empty.

And in general processing event timestamp minus output watermark doesn't say much about output data but more about backlog data.

I thought I understood it wrong but I see they make additional examples with data freshness spikes a bit lower in the doc and describe it:

"In the preceding image, the highlighted area shows a substantial difference between the event time and the output watermark time, indicating a slow operation."

Does anybody have a clear understanding of it?

Data freshness metric is basically a graph in Dataflow UI which shows the latency of your data in the pipeline. It very much correlates with system latency and is mainly is used as an indicator of performance and health of the pipeline. High spikes in the data freshness graph would indicate there is some delay (operations are slow) or any bottleneck in the pipeline. Which can increase the data backlog.<br><br>Output watermark is the processing time of a tranformation stage or stages in a Dataflow pipeline. A detailed explanation can be found in this talk given by a member of Dataflow Eng team.<br><br>One similar thread you can refer to is here.<br>