Thunderball,
I often haven't got a clue what I'm talking about.
A PhD in aerodynamics taught me exactly as you suggest - there is nothing
straightforward about aerodynamics and that I had more un-answered questions when I finished than I had when I'd started! They appear to be multiplying all the time.
The relationship I refer to merely illustrates that there is a fundamental difference between subsonic and supersonic flow behaviour in a converging diverging duct....which is essentially what an intake or nozzle is.
Depending on whether M^2 is <1 or >1, if you then look at the relationship between da/A and du/U, you will see something interesting.
In the case of supersonic flow, in order to decelerate the flow to a subsonic flow condition, you need to first, converge the nozzle, then diverge it.
Which is what a supersonic intake does.
If you now draw a quasi one dimensional duct of the former nature, and superimpose the pressure loads on it (they act normal to the surface correct?), then resolve them in the horizontal direction, you will see that its possible for an intake to generate a thrust.
I'm a great SR-71 enthusiast.
You know the flight manual is online?
http://www.sr-71.org/blackbird/manual/
