Sorry to be so late with this. I've actually been thinking hard about this idea after this study
came out. Basically, it argues that the constraint on humans isn't the brain ----> pelvic outlet problem, but the problem of overburdening the maternal metabolism. Which I think handles preeclampsia *better* than the traditional obstetric dilemma explanation, since brains are energetically demanding things to grow.
So yes, in preeclampsia there's a known U-shaped curve if you plot baby weights for gestational age at delivery. That curve seems to be the consequence of a dispute between maternal and paternal genes over how much energy is provided by the placenta to the baby. (The obstetric dilemma would say that's because the kid would grow too big to fit out if we let it have as many calories as the dad thinks would be a good idea; this new hypothesis would say that's because the maternal metabolic rate can't go that high - we just can't put any more energy into growing that baby.) All placentas have varying degrees of depth of implantation and all preeclamptic placentas have varying strategies that they are using to compensate for the shallow depth. (For example, a lot of us only put on water weight and postpartum drop below our starting weights, because the placenta has used the strategy of cannibalizing our muscle tissue during gestation.)
On this view, some of the offspring are small for gestational age - probably because the placenta was very shallowly implanted, and the strategy of ramping up our metabolisms still didn't supply enough nutrients and oxygen to push that offspring into the middle of the curve. Some of the offpsring are large for gestational age - probably because the placenta was able to compensate with these other strategies, the kid grew bigger than average, and we hit our metabolic rate limit and delivered.
One strategy available to the placenta when it's shallowly implanted is inducing gestational hypertension, and another is inducing gestational diabetes, so even if the two are related, the glucose test will still give you useful information. Some shallowly implanted placentas *don't* turn into hypertension, just IUGR - presumably in those cases the paternal strategies are blocked by the mother. We're still working on this, though, because it's what they call "multifactorial and heterogeneous" - lots and lots of different genes and strategies in play. The theoretical architecture is nice for explanations since we can't generate an entirely empirical proximate cause yet.
How are you doing now?