The difference is the incentive to improve, and actual present rate of improvement, for models like this is far higher than it is for jetpacks. (That and certain intrinsic features at least suggest the route to improvement is roughly "more of the same," vs "needs massive unknown breakthrough".)