Consider not running `*-as-if-foss` jobs in GitLab's `master` pipelines since GitLab FOSS pipelines would fail

A few quotes:

• Remy: I don’t like that we treat broken master for foss the same as other broken masters. This would make our success rate decrease for no advantage. Now that there are notifications into #master-broken we have a good process to tackle these failures. We can now open an MR for fixing it without the pressure for fixing it in the main project for heavy development.
• Albert: I do agree with that and we can use the gitlab-foss project as the single source for failures. Given #master-broken and the guidelines in the handbook we aren’t treating it as second citizen. My main concern is that if there’s difference between running in the master pipeline vs gitlab-foss project. If the effort to identify the problem is the same then there is no downside to running in foss.
• Kyle: How often are the foss pipelines run (looks like scheduled and every 3 hours for merge) https://gitlab.com/gitlab-org/gitlab-foss/-/commits/master

Benefits

• GitLab success rate won’t be impacted, while the work to fix the broken GitLab-FOSS master will be the same as today, with less pressure than fixing a GitLab broken master.

Drawback

• The failure would happen a few minutes later than now, since it would happen in GitLab-FOSS master after it’s been synced from GitLab.