If I had a infinite backlog of assistant tasks that generally generate more money then they cost I would. You're just leaving money on the table otherwise. If

If I had a infinite backlog of assistant tasks that generally generate more money then they cost I would. You're just leaving money on the table otherwise. If I had barely enough tasks to cover a full time assistant, or if assistant tasks were loss leaders, I'd probably go with the slow and cheap one though.