Yeah and it's super simple to roll significance and effect sizes in one with confidence (or credible) intervals. Plus the interpretation is super straight forw

Yeah and it's super simple to roll significance and effect sizes in one with confidence (or credible) intervals. Plus the interpretation is super straight forward: CI contains zero: Not significant. If nothing else, we should make CIs the primary default instead of p-values.