Run all the build steps regardless of the outcome of the puppeteer test

[davemfish]

Moving the puppeteer test to later in the sequence seems like the best approach here, since continue-on-error does not do what one would expect. We do not want a test failure to prevent us from deploying artifacts, etc.

Fixes #2454

Checklist

• Updated HISTORY.rst and link to any relevant issue (if these changes are user-facing)
• Updated the user's guide (if needed)
• Tested the Workbench UI (if relevant)

[claire-simpson]

Thanks @davemfish! My only question is should we keep the comment that the test is flaky? I guess my thought is that it might be nice to know that this test failing is potentially a GHA problem that we were previously able to somewhat ignore.

[davemfish]

Thanks @davemfish! My only question is should we keep the comment that the test is flaky? I guess my thought is that it might be nice to know that this test failing is potentially a GHA problem that we were previously able to somewhat ignore.

Yeah, I see your point. I don't think a comment in this location would be noticeable to someone exploring why the test failed though -- they just aren't likely to look at the workflow document. The comment was really only intended to justify the use of continue-on-error.