in a rough way,

AverageProfit = AverageProrfitPerTrade * ProbailityWinning - AverageLossPerTrade * ProbabilityLosing,

精准, if by it we mean to improve winning percentage, is a self evident argument. (The other story is to increase averge profit and reduce average loss, which is the classic "cut loss fast and let profit run". )

I can appreciate the statement that sometimes you should not pursue 精准 beyond everything else, for example, when you do a large size etc. But to qualitatively say you don't need 精准 without mentioning the context, it's really as helpful as it's meaningless.