Peer review of changes is still important.
Multi-model cross-review is important
You can't rely on people spotting the significance of such changes
Multi-model cross-review is important