Filtering approvers should follow more rigorous rules

[peternied]

Coming from https://github.com/opensearch-project/OpenSearch/actions/runs/7464605190/job/20311941616?pr=11490#step:3:8

• In the output it can be seen that there were multiple approvals from reta and peternied, they are being counted multiple times for the check of the number of required reviewers.

• The pull request reviews responses also includes changes requested and dismissed reviews, these should be be properly computed as removing previous approved entries. Or at the very least there should be an option to control this behavior.

filtered to approvers:
 ["reta","peternied","reta","peternied"]
Found approvals from reta, peternied, reta, peternied