Currently the functional tests are running with failFast activated, meaning that a feature will stop running its tests as soon as it has a failure.
This is important and useful in some contexts (e.g. PR checks) but does not provide all the information we may want since after a failure, some scenarios won't run and we may want to get that information.
What can also be interesting is to check if we can already integrate the bail flag on WDIO and Cucumber (see more), since we can use it to stop the entire run if more than X scenarios fail. (e.g. useful if some regression happens and we want to be notified asap)