Maybe im wrong, but it being a close loop placement of the pumps shouldnt matter
if it weren't for pressure drop...
so, yes and no.
As far as temps go, in the bigger scheme, the loop as a whole will equalise.. that does not mean you wont have fluctuations within the loop... the water exiting the GPU's will be (slightly) warmer as the water entering the GPU's, for example. wether this difference can be ignored or not depends on OCD values
But, as said by having the pump before a rad, you are not adding the "heat" created by the pump to the loop... overall maybe 1 - 2°C, if that.. but OCD rules
As for flow... closed loop, hydraulics at work, gravity doesn't matter (much).. as a whole there is no change.. but there are fluctuations.
Pressure wise.... you can add 3 D5's in series, creating a hug pressure front, then go through all rads, fittings, GPU's , then hitting the CPU before returning to the pumps. Depending on how much pressure drop throughout, your CPU-Block/Jet Plate might not be performing optimally. Not to mention the strain on the hoses/fittings just after the pumps
Also remember that a reservoir is a "zero pressure" zone. I'ld "evenly distribute" pumpt throughout the loop to mitigate the restrictions added .
Look at it like a rollercoaster... they don't launch you off with a 10G acceleration (nobody would survive that) to make sure it gets around the track...no, they have acceleration points at strategic points in the track... mostly before and/or after major "slow down" points.
PS: i know the rollercoaster is not the best metaphore
.