Diagnose the situation before picking the rule.
The right question, in my experience, isn't "should we move fast or get this right?" It is "what does this situation cost us if we are wrong?" Reputational damage, safety, regulatory compliance, irreversible commitments to customers. That is first-time-right territory, and the discipline to slow down there has been worth more, every time, than the speed I would have gained by skipping it. Product features, GTM experiments, internal processes, market hypotheses. That is usually fail-fast territory, and the discipline to ship before you are ready has mattered more than the polish you would add by waiting.