“agentic coding has measurably improved every velocity metric we track and, in some teams, driven an order-of-magnitude gains in output” Of course it has. But

“agentic coding has measurably improved every velocity metric we track and, in some teams, driven an order-of-magnitude gains in output” Of course it has. But is that a good enough measure of value? What about the business and customer outcomes from all the extra lines of code and closed JIRA tickets? Was it worth it?