start_epoch vs/or start_time for Firefox-internal metric measurements/calculations?

Should we use start_epoch or start_time from https://github.com/WPO-Foundation/wptagent/blob/a45466623a8a812cf389c1124d7ff2fd6c4ff8dc/internal/firefox.py#L1061-L1062 and https://github.com/WPO-Foundation/wptagent/blob/a45466623a8a812cf389c1124d7ff2fd6c4ff8dc/internal/firefox.py#L330, respectively?

Or, is sometimes fetchStart/navigationStart[1]?

There are cases where we need to get as close as possible to standard/Chrome timings, but due to other reasons, [2] have to match Raptor/do a one-off calculation**[3]**

[0] [wptagent] https://github.com/WPO-Foundation/wptagent/search?q=fetchStart&unscoped_q=fetchStart [1] [wptagent] https://github.com/WPO-Foundation/wptagent/search?q=navigationStart&unscoped_q=navigationStart [2] [custom-metrics] as in trying to calculate DOMContentFlushed from its raw timestamp timeToDomContentFlushed - fetchStart or other non-standard metrics, for which we have to be careful to use the right webpagetest timestamp / W3C * Timing API, etc. [3] https://github.com/mozilla/wpt-api/issues/97