The answer is that the time it takes the ball to travel vertically to the ceiling is much shorter for the people sitting on the bus than it is for the people standing on the sidewalk watching the ball in the bus traveling by. However, that means time and therefore verticle velocity is slowed for the people standing on the sidewalk to 3/5 relative to the people standing on the bus, hence the time-frame rates would be different for the different observers even though events would take place simultaneously.
Of course, in that case, to reject the "fact" that the ball travels at constant speed relative to all observers would be to reject the ball-relativity theory. However the ball-relativity theory may be wrong, I don't believe that the light-relativity theory to be wrong though as the math for light-speed relativity (below shows the relative reference frames for the measurement of doppler frequency, position, and time as well as the values for energy, momentum, and wave-number for quantum particles) is soundly based on the "fact" that lightspeed in a vacuum is constant relative to all observers.
In the bus/ball case above, "t" is the time it takes the ball to hit the ceiling relative to the bus (t=0.09 seconds = 0.09/3600 hours), "v" is the horizontal bus/ball velocity relative to the sidewalk observer (40 mph), "x" is the horizontal distance that the sidewalk observer travels relative to the bus reference frame (40*.09/3600 miles), "c" is the absolute speed of the ball (50 mph)...." t' " is thereby the time it takes the ball to hit the ceiling relative to the sidewalk observer (0.15 seconds)....this has an analogue to quantum decay rates whereby "t" is the rest decay rate and " t' " is the decay rate for a particle in relativistic motion.
This "fact" about Special Relativity has been repeatedly shown to work for the decay times of particles in atom smashers. The rejectors of Einstien's Special Relativity Theory therefore have a lot of explaning to do.
