The argument made here basically is that measured inside the inertial frame of the second firing, there's a quantity of kinetic energy c, which when viewed by an inertial frame moving at -v, obtains the value c-v. But after taking into account of the -v, the second inertial frame also obtains...