I need to do the math on this, but there's an interesting tradeoff between pop and ec growth here. First, I want to point out that the issue isn't maximization until your city reaches max pop, but rather of marginal benefit. If your city needs growth now, putting it at some theoretical "optimum" isn't always the best option if you are sacrificing growth to build commerce that can't be exploited.
If you build a farm now, you sacrifice the growing of cottages to towns. If you build cottages now, you sacrifice pop growth.
As others have pointed out, in the early game esp. you often need to stop growth -- but that doesn't mean that the best way is to grow your cities slowly! It could be that you should grow the city as fast as possible, then switch them from working farms to working cottages. For that reason, I think the earlier posts that suggest setting up diverse tiles that you can switch amongst may be a smart move, so that you can work cottages or mines when you don't want growth but switch to farms when you do. Following this logic, building farms on rivers and cottages inland makes sense, although then a Financial civ sacrifices several turns of bonus gold from the river tiles -- so maybe for them there should be more cottages than normal on rivers.