The author leaves the business needs completely off the table, even though it would help his case. If you have a limited amount of time to launch your product in your current round of funding, and you want a framework that helps a small number of tech employees build working first generation critical features reasonably quickly, handle a large number of tasks not critical to the company's value proposition reasonably well, can be maintained and scaled on a variety of PaaS options well long enough until the company is profitable enough to move to the next phase of tech infrastructure, is Rails attractive software? I'm not an expert with it, and I don't even enjoy using it, but I believe it is.