Well, a lot less scientific than the testing above, but I have some data that may be of interest. I was experiencing intermittent loss of fuel pressure in a Hirth F23 EFI powered (45HP high torque model) Moni. Pressure would drop from 42 to 2 psi. I had experienced this one time before over 50hrs (pressure came back on line pretty quickly that time).
But, I just redesigned the exhaust system and the fuel pump moved from a tunnel with ambient temps to the hot side of the firewall. Also, the plane had been in storage for 5 years, so I completely went though everything, replacing a lot. In the first couple of hours of test flying (over a 9k foot quiet airport) I experienced the loss of fuel pressure several more times as I went though Extensive troubleshooting. (note, the Moni is a "motorglider", and I've got almost 10k feet on an almost dead quiet airport).
After far too much time and money, not surprisingly, a simple answer- Too much restriction on the inlet side of the pump. The gerotor pump can't seem to tolerate much at all. The pump was cavitating/vapor locking. I had put that low on the troubleshooting list as this pump can supposedly lift 500mm, and I was level with the bottom of the tank.
I have an Andair duplex fuel selector, and used a fair amount of forged elbows and Parker hose as there just no space for traditional bent aluminum lines.
The Pierburg pump (100 lph rating) flows multiples more that the 25lph peak my engine needs. I couldn't find a decent quality smaller pump (which is maybe why Hirth uses it). 7.21287.53.0 PIERBURG)
https://www.kmotorshop.com/document/tecdoc/5/PI-0034_e_1599590.PDF
I didn't like the idea of pumpling 4x my peak flow, so I found a quality looking down convertor. Walbro says their gerotor pumps run fine down to 9v, so I plan on running at 9.5v (it runs smoothly well below 8v). Flow is about 80 lph at 9.5v, vs about 110 lph at 13.5v. I spoke with the mfg of the down converter in Utah, and they said it's designed to their specs (made in China), and they sell a lot to a couple of major manufactures. Very nice looking, vs the junk I found on Amazon.
https://www.powerstream.com/dcp.htm#2237
Most interesting of all (to me) is that the current draw of the pump dropped 50% once I eliminated the large (pump inlet side) restriction(and cavitation). The cavitation must be terribly inefficient. (I need to measure the flow again with no cavitation).
(note- gauge tee'd into the inlet side of the pump- misplaced my nice digital manometer, so blew the dust off of this gauge)