When operating in an uncertain environment there is huge pressure to stabilise the collective understanding of what is happening. Prematurely converging on an explanation can cause confirmation bias in interpreting new evidence (you only accept data that supports your view) and cause the org to miss weak signals because people move on from seeking an explanation.
Creating a human mechanism for noticing outliers (evidence that doesn’t fit the pattern) and asking “what other explanations are possible” is one way to guard against this when the stakes are high.