Dear Ben,
You are right on that we are talking about different things.
By "x" both Ian and me are referring to the position as used, for instance, in QED. This is the instantaneous position of "point particles" as described in a quantum field theoretic framework. There is not such a thing as "internal motion" for point particles. Moreover, the nonzero Dx (D is delta) is not associated to real motion of any kind (althouth I know that some few references claim otherwise for the Zitterbewegung).
If you want compute velocities/speeds you must use compatible positions and times. If "x" denotes position in QED, then "t" denotes time in QED and ratios as Dx/Dt and dx/dt are mathematically well-defined and with physical meaning. If by "x" you mean otherwise (as it seems that you mean with your "externally observable motion"), then you must also change "t" from that on QED to that in your own model.
Finally, I want to emphasize again that the Dx ~ (hbar/mc) written in my above message is valid only for a stationary electron in QED and that it is neither the De Broglie wavelength lambda nor the Einstein-de Broglie 'wavelength' postulated in 1924 by de Broglie in his unfounded mixing of non-relativistic wave quantum mechanics with special relativity.