My theory is that devs often trade pretense-now for research-later. This habit allows them to gain ~0.002 status points here and there. But now combine that with unknown-unknowns. The thing they're hearing—if they actually understood it—ought to cause them to abandon or completely rethink a change they're currently working on. But because their takeaway isn't new knowledge, but a placeholder for potential new knowledge, they end up losing 15 status points when it blows up, in a way that was utterly unforeseeable.