Yeah. Or after this wobble effect has died down. Given that the frequency of "wobble" must be >50Hz (complete guess), then for the first <1/50 of a second after it leaves the club, it might still, theoretically accelerate. But then for every oscillation after that, surely the forwards wobble would be counteracted by a backwards wobble 1/50th of a second later. At which point it deccelerates due to air resistance.
As for the balloon thing - I can see where you are coming from, but 99.9% of the balloon's kinetic energy must be due to the balloon accelerating off your hand (like when you flick an elastic band), rather than wobbles in the air?