Folks:
Object mass, not diameter (i.e. size), is often the parameter of choice in various ablation models of meteors and spacecraft . But these have to be calibrated using experiments. In the 1960’s there were several multistage rocket experiments where weighed samples of various materials were launched down into the atmosphere at various speeds. And of course there have been literally thousands of rocket and spacecraft entries that verify the masses at the heavy end of the scale. These were and are being followed up at small sizes by modified nuclear accelerator experiments in both the US and in Europe. Since the theories (as also the measurements) are based on mass, the actual size will depend on the density which can range from fluffy 0.1 gm/cc (Leonids) to rocky 3 gm/cc.(Geminids). How bright a given mass is when it becomes a meteor depends on its velocity. So here is where the math and numbers come in. Sorry.
A grain of sand (coarse) is 3 gm/cc and perhaps averaging 1 mm in diameter. The mass is then ~ 5 mg. A zero magnitude (Geminid) meteor is in the 1 gm range for an average velocity of 35 km/sec Conversion to magnitudes gives 8.3. For a slower velocity ~20 km/s, the value is 9.5 magnitude. For a Leonid, 72km/s the magnitude rises to only 6.7 . So a grain of sand is a size under estimate for visual meteors, but is about right for the average of radio scatter detected meteor. For visual meteors, one has to go to glass beads (6mm are used for jewelry, Marbles are an overestimate for average visual meteors. ) 3.6 mag at 20 km/sec., 2.4 mag @ 35 km/s, 0.9 mag @ 72 km/s. For Leonids, the velocity does a lot. A grain of sand mass, 6.5 mm diameter “dust bunny” gives 4.3 mag. For fireballs, the mass goes up from marbles.