I avoided putting epoxy on top, but never accounted for the wax/grease being thrown out by rotation.
Just to have an idea about what is happening there, the rotation speed (rpm) can be calculated (and implicitly, the centrifugal force). In those calculations (in case anybody is curious) it's up to everybody to assume the velocity and the rate of twist.
In those conditions, one needs something much stronger than wax (as a plug closing the cavity) to make sure it holds together. But wax (or other soft materials) might do for slow things, 22 rimfire, HG cartridges, etc.
Nevertheless, a "plug" or an opening of a SMALL diameter will only help things.
Also, the hole is supposed to be thin and long as opposed to large and shallow for two reasons: minimizing the radius (and by doing so, minimizing the centrifugal force), and for obtaining a large amount of pressure concentrated on a small surface during impact.
At least, that's the theory of it.
The idea with the mercury or glycerin is at least 90 yrs old, and I wondered nobody mentioned it in here before me. Modern day experiments included elastic materials, like some hydronic types of rubber, but apparently they didn't work as good. "Compressible" oils was another hope in the 50's and 60's, but I don't know about any fantastic results.
Again, for information only.