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.