I think the difference vs. taskrabbit is that it being realtime requires the task-runners/execs to be higher quality and more committed, so it might not work as well in a market with 0% unemployment where everyone has a 9-5 job. I don't think that's a major risk for the next decade or two.
They can obviously raise pricing. At some point, people will pay based on the money they save, and runners might get really efficient at certain kinds of missions. If, for instance, a common mission is "pick up my car from the bar the next morning", the requestor might think it is a 2 hour task (bus or taxi, ...); maybe a task-runner has an efficient solution involving a folding bicycle, or two people. Worth $100 to me, costs 30 minutes of someone else's time who is happy with $25/hr, win/win at $50.