I think this is probably the chapter that GIT is talking about, it's about Mach's Principle in a book edited by Barbour and Pfister:
http://books.google.ca/books?id=fKgQ9Yp ... te&f=false
GIT, I'm not sure that anywhere Chris' analysis requires a rotating universe. Yes, he makes an analogy to a car moving on Earth, but I think the main focus of the analogy is that so long as you have (1) a power source that converts fuel to energy and (2) something else to push against, you can move while conserving momentum without needing to eject propellant (ie. you do not need to carry your own reaction mass).
In the case of this analogy, a tiny bit of momentum is transferred from the car to the Earth, altering the Earth's speed of rotation by a miniscule amount, although when the car brakes momentum will be transferred back to the Earth (after all, deceleration is equivalent to acceleration in reverse... with a conventional rocket you have to expel propellant to slow down as well as to speed up). But, it's just a partial analogy. I don't think the rotation element was intended to carry over.
In all the discussions about this I've been under the impression that most of us think that momentum has to be conserved by being transferred to Far Off Active Mass (FOAM) or, as it's been called in this thread, the Rest Of The Universe (ROTU). People here have repeatedly made the joke that, as a result, it might be the mass use of Mach Effect ships by alien spacecraft across many galaxies that is causing the expansion of the universe to accelerate instead of slowing and then collapsing back into a Big Crunch.
The more general problem we seem to be back at is "how does the M-E device know how fast it's going compared to the rest of the universe?" Some have argued that relativity requires that no frame should be preferred, and that then the M-E thruster should always get the same thrust in each instantaneous rest frame, so that it will produce constant thrust indefinitely. Then we get the rocket on a flywheel argument, and the disagreement between people who say that Newtonian math will come to the conclusion that any rocket can eventually go overunity by producing constant thrust and that you have to do Lorentz transforms to get a non-absurd result, and people who do the Newtonian math and conclude that Newtonian math does not get overunity for any type of rocket.
I can't do the Lorentz transforms and I can't find any obvious errors in either Andrew Palfreyman's or Chris' math regarding rockets - but if there are errors I wouldn't necessarily catch them. So most of us are stumbling around half-blind in this discussion due to lack of math skills. What I am
interested in is following detailed discussions in order to try to learn something and see if any criticisms make sense to me.
So, going back to Chris' analogy. Does the universe have a rotational angular momentum? No (at least assuming Mach's Principle is true). Does the universe have a momentum in a certain direction? No. However, mass in the universe does have a momentum - it's expanding outwards because some expansionary force is stronger than gravity. So, you can perhaps in theory exchange momentum with that large amount of far off mass which is going away from you, in the same way that you can exchange momentum with the planet you are standing on (notwithstanding that the rotation element is not analogous).
From this point of view FOAM is probably a better term than ROTU, because it reduces this confusion. Nonetheless, it seems to me a bit of a semantic difference: the thruster and the FOAM it reacts with are all part of the universe, which is a closed system which conserves momentum (if we agree with Woodward, then the FOAM is the vast majority of distant matter in the universe, although I've once seen the argument made that it can only be the far off matter within a certain prescribed light cone (but that's a different argument and doesn't necessarily affect the outcome here, as it might still be possible to treat all parts of the universe within a given light cone as a closed system).
We also get back to the argument we were having involving Chris a few weeks ago about whether the reaction on the FOAM should accelerate it or decelerate it. I tend to think that even though it's already accelerating away from you, you still accelerate it even more by exchanging momentum with it, similarly to how you always push propellant away from you to speed up or slow down. But, I might have a conceptual blind spot here. In fact, I'm even finding it hard to conceive what sort of momentum exchange you have with Earth when you drive your car on it. Would you be slightly accelerating its rotation if you drive east (against the direction of rotation) and slightly slowing its rotation if you drive west (in the direction of rotation), or slightly increasing its polar wobble if you drive north or south?
And yes, "push" might not be the right word/concept either, given the whole gravinertial transistor concept. But still, in Woodward's theory it seems to me that when you take advantage of gravinertial flux you still transfer momentum by doing so to the FOAM. So, for a discussion at the level we're having here, "push" seems to be an acceptable analogy to me.
Overall, though, I would find it interesting if, simply using Newtonian analysis and "high school" level math, we could reach an approximation that it is indeed possible in theory to push against FOAM (assuming there is a mechanism to do so like gravinertial flux/Wheeler-Feynman gravity radiation, or whatever someone else might propose) while conserving momentum (and not get constant thrust but actually need more energy to keep going faster relative to your original reference frame.) Such a Newtonian approximation may not describe the behavior of a propellantless thruster entirely accurately, and the higher the accelerations/speeds and the larger the distances, the more it might go completely off to left field as compared to a model using GR and SR... but, if on smaller scales a Newtonian analysis provides an OK approximation it would seem to me to actually be favorable towards the whole concept of propellantless thrusters.
Perhaps Paul would like to try to comment on or clarify some of the points above?