It would be valuable to read your spin on these from running VWO and seeing A/B tests for the last X years. For example, should I bother putting in images from

It would be valuable to read your spin on these from running VWO and seeing A/B tests for the last X years. For example, should I bother putting in images from different angles? Are they any demonstrable advantages of doing so. Appreciate that you should test pretty much everything, but some of these seem superfluous.