I just did my own version of a "myth-busters" experiment in the kitchen. (Sorry, no explosions)
I took a small saucepan about 200mm diameter and measured 500 ml of water and placed the water into the saucepan.
I allowed the water to sit for a couple of minutes to stop any movement, then placed a single drop of food colouring into the middle of the pan.
The saucpan was then fairly gently moved from side to side, and the food colouring widened its footprint to about 2/3 of the pan width after 6 side-to-side movements. Only minimal mixing with the water was observed.
Then the experiment was repeated with only 100ml of water in the pan.
After the same degree of side-to-side movement, the food colouring had not only spread over the entire bottom of the pan - it was pretty well mixed through the water as well.
Experiment - "CONFIRMED"
Different car i know being a Nissan TD4.2 but because i fiddle and the fuel pump which is adjustable in just 5 minutes, i have had lots of plays with it. Upshot is even stock diesels are quite variable in the amount of fuel they can be tuned to use. That little often spinning gizmo inside has no concept of fuel volume or air volume but only centrifugal force (rpm) and this device in this dumb sense allows the appropriate volume of fuel to be used as judged only by this parameter.
Turning down the fuel or for the non-adjustable types the original setting (hopefully within spec) can make for a good degree of difference in performance because in a diesel as opposed to a petrol more fuel will always = more power and if no more power is readily possible then extra heat, if the pump is adjusted to run slightly rich in comparison at highway speeds it will and that will be reflected no matter how the person drives it.
I can vary my cars highway fuel consumption by up to around 10% with nothing more than a twiddle of it's pump, i cannot make as readily measurable difference to town km with the pump and with the same sized adjustments, I think this is understandable.
Point is it's often not the engine it's self but almost always the setup however that is defined and 10% is entirely realistic between identical diesels and still be drivable but possibly not preferable.
I sometimes add ULP to my diesel fuel at around 5% by volume then turn the pump down an extra amount...when I do this I have without doubt the fuel efficient TD4.2 patrol wagon in the world still while keeping more or less it's genuine performance.
cheers fnq