[b]GN:

What you're getting is not reasonable, unless there is another problem.

With the engine on your stand disconnect the pump, loop a short length of tubing from the pump inlet to the outlet, plumb the fuel tank directly to the carb bypassing the pump completely. Run the engine and get it stabilized, any particular rpm, doesn't matter. Now undo the looped hose. Any difference in the engine run? While it's still running plug the pump outlet. A difference now? Still running, with the pump outlet plugged attach a hose from another tank of fuel to the pump inlet. Did it suck the fuel up? Unplug the outlet until fuel comes out, then plug it again. Any difference in the engine run yet?

Now, in case you wondered, all this is to check for an internal leak in the pump, and I'm sure you've already realized that. Any change in the engine run while doing these tests to the pump will show a leaky diaphragm.

As a final check, switch first the carbs between the good running and problem engines, then the pumps. If the pump passed the leakage test the further switching will show if the pump or the carb is the problem, if it persists with both the carb and pump from the good engine then it's in the basic engine. If the problem switches with one part or the other, you have at least found where the problem lies.

HTH.

Bill.