The carb realy doesn't care about the pressure as long as its enough to deliver the volume getting emptied from the bowl. Higher pressure only increases aeration, and too high of pressure can force the needle valve open - especially when idling and coasting. Steady pressure in the 3 to 5 psi range is best because it will provide the most consistant fuel level in the bowl.
A return line can help keep fuel cooler than if its sitting in the line over hot pavement or in a hot engine compartment.
It can also make it easier to control pressure, especially with lelectric pumps using a high pressures needed for FI.
Mechanical pumps increase volume and mass flow with rpm.
Vapor. When gasoline vaporizes the pressure goes up.
The description you have posted is that there is pressure at idle - which is the condition where the engine is consuming the least amount of fuel.
The fuel bowl inlet valves are closed or nearly closed.
The reason the pressure drops with increased throttle must relate to a lack of volumetric flow. The suggestions to figure out whether this is due to a restriction or a pump failure is the way to solve this.
Ideas to persue
1. Hysteric or someone mentioned electric connections. If the pump is connected to a circuit that is a long distance and/or sharing power from the alternator, then there could be a drop in voltage and therefore poorer pump perfromance.
2. Location of the pressure measurements is critical to figuring out what's going on. Measure the pressure closer to the pump outlet, then the regulator outlet.
3. Do a set of pressure vs volume experiments. Skim through this, and then go back and read through the posts there as needed to understand. See what cudafever did to prove it and then I think you can apply it to your situation.
Holley/Demon Carb stumble (posts 472 - 491) See in particular the sketches of Cudafever's system in post 485, and his tests (488-491) that show lower pressure allows more volume to the carb.