Here is what we know from bag testing, which is as close to real gph as you can get, it measures what can actually be collected from the pump in a bag, no computations, no guessing, no theory, turbulence is not measured, only what can actually be collected. We used the following method, a pipe of 10cm length with a diameter matching the pumps output, with what is essentially a heavy duty trash bag on the end, it is deflated of any air and held closed by one hand at the pipe until "go". The pump is running in a tank about 12cm (5") from the surface, one person attaches the pipe at the word "go" from a stop watch reader (2nd person) and the reader counts down 5 seconds and says "stop" at 5 seconds. The bag is weighed in kg after being placed in a rigid tub to avoid tearing and spills, the scale is tared for the weight of the tub and bag empty. 1 liter of water weighs 1 kg, there are 720 5 second intervals in an hour, by multiplying the collected weight of water in kg by 720 you get lph. We do 5 runs rotating the watch reader and the bag tester to minimize reaction time errors (though a person who is slow will generally attach slow and remove slow and same of someone who is fast so this error is largely self correcting) and we take the average of the 5 runs. Keep in mind that the bag and pipe are a source of resistance and in all likelihood, real numbers are higher, but not astronomically, likely no more than 10%. Anyone can do this test themselves. To convert to gallons, simply divide by 3.785.
6105 with 18V jumper-9620 lph (Average of 13.36 kg of water collected in 5 seconds)
MP-40- 11433 lph (Average 15.88 kg of water collected in 5 sec)
6105 with 24V jumper-11950 lph (Average of 17.48 kg of water collected in 5 seconds)
You can also see why this method is so limited, much past 20kg and you will just end up soaked when the bag rips open. So, bigger pumps and longer sample times which would increase accuracy become very difficult
Thank you for the data.