Reinstate automatic retries in review-qa-* jobs

What does this MR do and why?

Describe in detail what your merge request does and why.

Since we removed automatic retries in review-qa-smoke and review-qa-reliable jobs in !87038 (merged) we have been observing an increased amount of TCP connection errors that appear to be caused by race conditions such as k8s node reboot, see #362980 (closed). It appears the failures are intermittent and retrying the job manually in those cases helped to get a green pipeline, so I suggest reinstating an automatic retry while we are investigating. This should help to prevent merges being delayed in case of a failure during MWPS.

MR acceptance checklist

This checklist encourages us to confirm any changes have been analyzed to reduce risks in quality, performance, reliability, security, and maintainability.

• I have evaluated the MR acceptance checklist for this MR.