But not trying something very unlikely to succeed sounds valuable to me. Why waste 2 months of engineering time if your engineering team is convinced it will take at least 6 months?