I disagree with the reverse tadpole design having superior stability for one specific reason: People don't plan to have to hammer the brakes and swerve in an emergency, an action that often generates the highest g-forces to which the reverse design would still be more susceptible to instability and or rollover with all else being equal.
In more technical terms, the distance of the center of gravity to the fulcrum line between the front and rear tires is much shorter with a reverse tadpole under hard braking and turning than it is from the CG to the contact patch of the front wheels on a standard tadpole.
These are forces that one is unlikely to be able to replicate to the same degree under acceleration and would be more expected/planned for as well.
If the theoretical tip-over limit exceeds the maximum g forces the tires can produce, it doesn't matter what you do, it won't rollover. Unless you hit a curb going sideways or something, but that would roll even a Corvette. "With all else being equal" is correct, but only an idiot would put the engine in the front of a reverse tadpole trike. AKA, everything you said I agree with 100% wrt to one of those stupid and dangerous Goldwing trikes.
If you take a look at the article I linked, the designer talks about the stability of the Vigillante, which is an example of a reverse tadpole "done right". He distributed the weight so that under full braking the weight distribution was 33/66 fr/rr leading to a stable condition, even under full wheel lock. Impossible to roll, due to the low/rearward CG, and only able to lift the front wheel very briefly under hard acceleration from a stop. Its very low polar moments of inertia minimized the side-load needed on the front tire for turning.
Thinking about it further, I suppose it could be said that a tadpole "all else being equal" would have better braking/turning stability, but it's really a moot point if either design has a theoretical tip over limit exceeding the tires maximum grip, since in the absence of hitting something that would tip any wheeled vehicle, rollover is not a concern.
As far as performance, under braking a tadpole that distributes 66/33 fr/rr vs. a reverse tadpole that distributed 33/66 fr/rr . . . both should have equal stability.
Under acceleration, the reverse tadpole wins, as it can distribute nearly all of it's weight on the rear wheels. A FWD tadpole would shift the weight away from the drive wheels, and a RWD tadpole that maintained 66/33 fr/rr under braking would never have the traction the reverse tadpole would have.
All great points. That car illustrates very well that despite the theoretical limitations of the reverse tadpole, it is effectively relegated to being a non-issue. On that point, I was clearly splitting hairs.
Your idea for using a Geo metro drive train is a good one I have been thinking of doing this too. I looked at the Geo metro and it would be easy to drop out the front suspension and drive train complete with steering column. build a frame for a trike around it, add a very light highly Aerodynamic body and 100 MPG should be do able