More on "no-fall" tops.
The simplest approach is a top with a thin "platter" flywheel which has a Cg lower than the center of the (large) ball tip. I've made many of these. I even have a spreadshseet which calculates the height of the Cg. Example, a top with 0.5" diameter ball and Cg lower than 0.25".
In addition to this, I've found that a very well balanced top won't fall if the Cg is only slightly above the center of the ball.
Finally, getting the Cg low, even if not low enough to prevent fall, might provide extended spin time because of low topple speed. I made a 1.75" diameter, 54g top tonight which topples at 85 RPM, less than half the topple RPM of most of my small tops.
But there is "no free lunch". These tops have larger balls and lower scrape angles. The greater decay rate of the larger ball consumes the benefit of the lower topple speed. And, do we enjoy 25 minutes spins if the top scrapes half the time we twirl due to the low scrape angle?
Developments full of trade-offs like this consume a lot of time. I'm constantly trading parameters in pursuit of best compromises. And best for me, won't necessarily be best for other twirlers.
Alan
PS I haven't forgotten the original topic of this thread "A Figure of Merit for Twirler Spin Time". I compute merit on every top I make and twirl. I find it helpful when comparing tops of different size and weight. A high figure of merit on a 50 gram top says, "This is a good performer, considering it's 50 gram weight". Conversely a low figure of merit tells me, "This design needs improving", or "This design is inferior to that (other) design". etc