[billbo911]

Originally Posted by Nuttyman54

The problem is that the reductions you get from those stages are to small to realistically do a direct chain reduction, so you'd need to add an additional gear stage...at which point you're back at something akin to the available COTS options. You're trading reduction for spread

The stages the OP has modeled are 18:60 and 24:54, which are 3.33 and 2.25 reductions, respectively.

The alternative gearing you propose to get the 2.27 spread (24:60 and 40:44) only give 2.5 and 1.1 reductions, respectively.

You could maybe do a 34:44 reduction instead of the 24:54 reduction, for a 2.57 spread (almost the same as an AndyMark Supershifter), but that still ends up awfully fast (20fps range) with a 12:36 chain reduction.

No doubt about that, it is NOT an ideal way to go. There needs to be a more reduction at the first stage to make this a viable solution. I was just trying to get the spread a bit greater.
Honestly, after using the original ball shifter for two years now, we would like to see a bit more spread than 2.27.
2.57 is probably closer to where we want to be, although even that may not be absolutely ideal either.
Ahhhh compromise, the bane of all things Engineering!