The only way in which the power factor n (in the air resistance R=b*v^n)
does not matter very much is when b is very small. Suppose that b is so
small that R << m*g at all v from the experiment. But in that case the
monotonic decrease of a with time must be attributed to something else,
perhaps the artifact of "averaging". Does anybody has access to ballistic
tables? It would be useful to know how large is R for a sphere whose
tables? It would be useful to know how large is R for a sphere whose
diameter is 9 inches at v=2.5 m/s and v=5 m/s.