Really, Peter, these discussions remind me of a poem I read as a wee lad:
http://www.noogenesis.com/pineapple/blind_men_elephant.html
The load changes as the angles change.
Well, that's true of course, but for a great many systems IT DOESN'T MATTER! Unlike the encounter with the flying cow, the change in angle and load are very gradual. For all practical purposes the load change will be compensated by a gradual increase in pressure. Technically there will be a slight slowing as the pressure increase causes expansion of the cylinder and hose, but that effect is going to be negligible and imperceptible in all but a very specialized system - perhaps motion control. Back in the day, when I did field service on constructiom machinery, if I had detectible softness in a hydraulic system I knew it was time to bleed system and get rid of the entrained air.
No body cares on a system like this but when the same techniques are used to design motion control systems there is often trouble.
And that, sir, is exactly my point! The fact that an engineer recognizes the difference between a servo system and construction machinery, for example, isn't a sign of incompetence. It means that the analyis is appropriate to the system. There isn't any point in doing an analysis of a motion control system when you are designing a crane, for example, or an elevator. Adding inconsequential terms to an equation or including consideration of negligible factors doesn't add value to the final product or indicate superior analysis. They just waste valuble engineering time.
You are assuming the pump is always pushing oil over the relief valve when not moving. That is not energy efficient.
But, of course, a gear pump IS always pushing oil. If the spool is in center position the flow is pumped right back to the sump at essentially zero pressure so there is no pressure drop or power. If the pump path to the sump is blocked, say if the spool is shifted to direct oil to the cylinder, then the oil must either flow into and move the cylinder or dump over relief!
The energy inefficiency may or may not be correct. It really is a function of duty cycle and system operation.
Pump hp = psi x gpm / 1714
A gear pump in an open center system dumping full flow at 0 psi draws 0 hp. A pressure compensated variable displacement pump in a closed center system at zero flow and max psi also draws 0 hp. During idle portions of the cycle, then, it's a wash. A lot of systems spend a lot of time at idle.
If an operator is pushing production he has the control valves at max position. A gear pump pushes its constant flow rate against the actual load pressure. A pressure compensated pump will stroke out and push max flow against actual load pressure. In this operation power draw in the two systems is equal.
The power advantage of a closed loop system only appears when the control valves are modulated and the system is operating at reduced flow. A gear pump will dump the balance of the flow to the sump at full pressure, wasting power. A pressure compensated pump will stroke back, reducing flow and saving power. If the hydraulic system being designed will spend a lot of time in this condition then the energy savings may offset the higher cost of the more sophisticated pump. A lot of construction machinery uses plain old gear pumps and open center valves. Many systems use multiple pumps so power isn't wasted when a particular circuit is idle.
And, for a lot of systems (like a backhoe) the flows are so high that sizing an accumulator to do more than attenuate pressure spikes just isn't feasible.
There are many 'logic' valves and other unnecessary junk in the system.
That doesn't sound like confusion over the role of pressure and flow to me - it sounds like the inability to incorporate the KISS principle and failure to understant the actual physical constraints of the system.