On the pump the MPT inlet especially if your using a coupling the fluid flow will go from .875" to .455" at the flat end of the pump inlet, this will create a large turbulence point. My suggestion is to countersink the inlet ID until the outside of the CS is just touching the last of the faced off inlet surface for a smoother fluid transition. Every turn, bend and fitting adds turbulence to the flow not counting tubing friction, it all adds up to reduced pump performance output.
I just got inside my pump, the side inlet port from the housing to impeller is "D" shaped, width of .890", the impeller cast center post .490" diameter.
The pisser my pump has stainless flashing at the end of this center post like a mushroom a diameter of .620". This alone is a grain and crap snag location besides causing turbulence as well restricting impeller inlet flow.
Looking into the inlet port there is a ramp with a step that also that can be smoothed down for a smoother fluid transition, possible reduction of pump harder to get primed is my thinking on the above mentioned. The side of the impeller to inlet port is a rather wide gap causing fluid to bleed past vs caught in the impeller and centrifuged out to the discharge port. Again my thinking here to up the pump's efficiency from the standard production design, it all adds up and can only help.
Adding to this there is quite a lot of impeller end play which increases bleeding across the input port vs shearing and pumping out, my pump this gap is .140" wide when the impeller is against the stainless cover plate.
I also found I was able to slide the motor magnet out an additional .094" just clearing the stainless cover, this allows the impeller magnet to just float touching the inner side of this stainless end cover reducing the pulling of the two magnets when they are not inline magnetically. I see this as less end thrust pulling wear plus deeper engagement of the drive to driven magnet in strength for a stronger impeller pulling before the magnetic coupling drive starts to slip.
I have a wild idea, narrow down both sides of the housing for less fluid bleeding down off the sides of the impeller blades making the pump more like a positive displacement in function plus chamfering and blending the inlet and outlet ports. Later on i'll first do NIB flow numbers then redo testing after reworking the pump to see if any actual number differences. I must move around back's killing me just typing this.....=o&o>........