Thom,
this is leaving my area of expertise...but since you mentioned me, I feel compelled to answer anyways!
I seem to recall reading about something like the kind of injection steering that Dave pointed out.
I believe it was mentioned in what seems to be a well-known rocketry textbook, "Rocket Propulsion Elements" by Sutton.
It was merely about altering the amount or possibly the mixture of fuel and oxidizer, and I forgot what rocket it was used in, but the basic engineering principle seemed to be that in a length-optimized chamber-throat-nozzle arrangement, mixing of the propellants from far-away points on the injection plate is rather incomplete, so differential injection leads to differential thrust.
Using a larger range of conditions along the chamber and nozzle wall, where heat transfer is greatly influenced by presence or absence of a sufficient insulating boundary layer of gas, will obviously be bad news for the guy who has to (experimentally?) validate the rocket's reliability. Otherwise, I think it is a very clever approach to make do with few (and maybe even only two, if spinning the rocket?) control valves for steering, potentially wasting far less propellant than in a cold gas thruster, and potentially the most weight-saving approach (blatantly assuming that no kind of motorized steering will be lighter than a valve).
Just let me stress again: This is less than 2 cents worth. Please check the source, which, shame on me, I did not do for this post.
Sincerely,
today's one-cent-worth pyramids