I notice two different plumbing arrangements in the examples. Bob's photo shows the pumps plumbed in series (outflow from pump #1 goes to inflow for pump #2) while the fittings for the other photo looks like it's set up for parallel plumbing (inflow from tank is split going to both pumps and outflow from both pumps comes together on it's way to the motor). Seems like both would work fine ..........................
Well, I feel the "proper" installation for function is to have them in parallel, as each unit has independent access to the reservoir and is adding to the total sum of delivery; versus a series installation with the singular draw sum effecting from only the first unit, which is the weak side of the system, with the potential gain solely being the perhaps drop in discharge pressure of this first unit due to the attempted draw of the second, and increase of the inlet pressure of the second, which one would anticipate an overall increase in delivery, but not at all equivalent to the parallel installation.
Also, consider the effects in the event of failure of one unit?
And as being opinionated, I would comment that in the photo of the parallel installation plumbing, in that the choice of the "T"s as somewhat lacking, as particularly on the low-pressure side the combination of the fuel turning 90 degrees into the first pump, over the machined fittings' short-turn creating a shear and the separation of the volume to the second pump unit simultaneously may induce a pressure drop, thereby an increase in the potential of a cavitation event, this definitely effecting delivery performance negatively. Now, this may be offset by utilizing plumbing of excessive dimension, but this just compounds in inefficiency of the design.
Scott.