One quick option is to use the hit depth dimension with your event, this will show you the distribution of hits that happened before your event.  Another option would be to use Audrey Salerno's great tip from Adobe summit.  https://www.dropbox.com/s/fnxsr9s94k63rsk/RockstarTips2019_AudreySalerno.pdf...

You can use an if statement and the percentile function to do something like 'if pageviews > 99th percentile, then average view time, else 0'.  Then if you sort the table by view time, it will effectively only sort those above the 99% criteria, everything is will be zero.

If users attempt to create an account with an email address already assigned to one, do you capture that error message within analytics?  If so you could exclude visits with that error message and visits where users select 'log in' from the registration page to do your analysis.

Have you stored the UTM in an eVar or in the campaign variable?

That link appears to be broken.

A visit with a single hit will not have a recorded time.  This is similar to how anything that happens after the last hit, before the user departs the site, is not included in time calculations.

Within the context of Adobe, 'data warehouse' and 'data feeds' are two distinct things.  Within the data feed, each row represents all of the data collected in a single analytics server call.  That table is supported by some look-up files, as referenced here:https://docs.adobe.com/content/help/en/an...

At a percent and volume level, how different are the values?  There may be very slight differences depending on the API used as well as which metric you use within data warehouse.https://helpx.adobe.com/analytics/kb/comparing-visits-all-visitors.html

This already exists as 'Return Frequency'https://docs.adobe.com/content/help/en/analytics/components/variables/dimensions-reports/reports-return-frequency.html

Thank you for checking in Justin.