Saturday I went out and remileaged my rally. One of the instructions has a speed change that could occur at two different points within the same intersection, and I remember deciding on one early in the planning stages. But I couldn't remember which point I'd decided on. So I took down the mileages for both, and figured I'd decide over again and use that one.
Comparing those mileages to the ones I took the first time, though, I found that the repeatability on my computer was quite excellent -- although they are only 50 feet apart I can easily tell which execution point I had chosen the first time around. And in general the standard deviation of the wobble in my mileage measurement is about two feet. Vastly better than GPS, even leaving aside the sample rate problem that GPS has.
So I think perhaps I should name the prototype model h0. After all, the Hubble constant expressed in seconds is a pretty tiny number.