Q1. How much single-twirl launch speed would you lose with substantially more AMI?
Q2. Could you get any of it back with a different stem profile?
Q3. Could you reduce the CM-contact distance even more?
Q1: Difficult to say.
Q2: The stem is a bit short, with a longer one I believe that the top could be started at maybe 1200 RPM, instead of 1000, and spin for maybe one minute longer.
Q3: Yes, if I reduce the topple speed from 150 to 50 RPM the spin time could be maybe 10 minutes longer. And the top could spin for maybe 28-29 minutes, with a single twirl, (starting from about 1000 RPM). Which is not, even so, extraordinary, for a top with the tip recessed. My best tops in this category spin for 45 minutes, (50 minutes with the lead flywheel), with a single twirl.
I am not sure about the effect of making the flywheel even thinner and heavier, with more brass.
The top we are talking about is the Nr. 33, in the graph below, (about the percentage of RPM retained after each minute of spinning).
It is very different from my other tops;
it is by far the worst one at high speed, losing 20% of the RPM after just one minute spinning, when the other tops lose only 4-10 % at the same speed. Too much air drag for that AMI.
But that same top becomes my best one, below 200 RPM. The air drag is no more a problem, and having the best AMI/weight ratio makes it the most efficient at that point.
I know that you remember, Ortwin made some extraordinary tops, made like a space station, in some way similar to the Nr. 33, (as for the very good AMI/weight ratio), but using spokes instead of a full disk, so not only he had excellent efficiency at low speed, but also not too much air drag at high speed, and he had some fantastic single twirl spin times.