[Aren Siekmeier]

I agree that the whole "sig figs" deal is flawed. I had a science teacher in high school who extolled it, and perhaps it was his inability to do basic algebra that led me to skepticism of that as well. My beef with it is that it is so dependent on the base in which you are representing the numbers. If you're going to worry about such things, actually do the stats with how the initial standard deviations or uncertainties propagate through the calculations. If you don't care that much, just round your final answer at the end to something that's not way than you care about, but still gives detail relevant to the problem and its context. My answer 10.43 was perhaps too many digits: 10.4V would perhaps have been more appropriate given how much control we actually have over (and how much we actually about) the voltages in question. But it doesn't really matter that much. And as Ether has pointed out, 11V is just plain wrong. And one should never round in the intermediate steps. The best way to go about these calculations is to solve for the answer symbolically and then just plug in your parameters, since the general solution in infinitely more useful and this avoids any intermediate rounding errors. I do a lot of calculations in excel and keep all the digits (the first 30 some anyway) around. However, when reporting these intermediate values, I will often round to keep it uncluttered.

Also (to sanddrag mostly), my method of solving was based on the fact that stall torque and free speed scaled with voltage, so I whipped a quick mathematical model that did just that. Your (sanddrag's) approach went through the physics of it from the start, which is equally valid (in fact, that's where we get the linear dependence of stall torque and free speed on voltage, as well the linear performance curve between them).