was there a rule against them going below the point? if the post is only 2cm wide, i'd have put the moment of inertia ring down below the balancing point. i'd also use a kinitmatically perfect ball as my point pressed into a socket (looks like thats what the aerospace guys did), and insist on the surface not only being ground but scraped/stoned flat. rather than your balance problem getting worse as its sways, having the ring below the point will auto-correct that.
EDIT: opp they explain why a ball is the ideal tip. but they miss the most important part .... concentrically of the tip. and instead of glue, mount it with a draw-bar style screw.