How many more weeks in utero does Karp think would be ideal? I don't know Karp at all, but a quick search indicates he's known for soothing babies by swaddling. Great. My mother swaddled all her infants many decades ago.
If he is comparing human neonates to primate neonates, he's correct. Baby primates are born pretty much knowing how to cling onto the mother with four extremities, necessary because primates are quadripedal, and need their arms to walk as well, meaning baby needs to hold on, aided by opposable toes. (Chimps and gorillas do support the baby with one hand longer, though.
But what evolutionary advantage would it serve for our infants to come out grasping? humans are bipedal, meaning we have arms that can be used to carry an infant. If we need our arms, we have the intelligence to fashion slings or carriers.
The same goes for any ability: sight, motor skills, etc. They don't need to be born with these skills, because they are cared for.
Antelopes drop a newborn which is up in minutes and able to see and run. Cats drop a newborn which can wiggle, crawl some, is blind and as helpless as a human.
Babies, through evolution, are able to live outside the womb (assuming no complications) without extraordinary maternal assistance from about 36 weeks on. A look at growth charts show steep rises in both cranial size and weight gain.
But you can see that a baby's head at the 10%ile takes at least six weeks to catch up to a baby the size of one born at 90%ile. Why doesn't that baby just gestate for 6 more weeks?
There is much more than head size going on here. There is the placental size, the amount of amniotic fluid, the possibility of meconium aspiration, etc.
Stillbirths are twice as common at 42 weeks and 4 times as common at 43 weeks.
Approximately 20% of postterm (41 weeks gestation) fetuses have fetal postmaturity syndrome, with chronic intrauterine growth restriction from uteroplacental insufficiency. These pregnancies are at increased risk of umbilical cord compression from oligohydramnios, [low Apgar scores], intrauterine passage of meconium, and short-term neonatal complications (such as hypoglycemia, seizures, and respiratory insufficiency). ...[T]he 4-fold decrease in the incidence of the meconium aspiration syndrome in the United States from 1990-1998 has been attributed primarily to a reduction in the postterm delivery rate [(induced delivery at 41.5 weeks)]. Postterm pregnancy is also an independent risk factor for neonatal encephalopathy and for death in the first year of life.[17, 18]
Maternal complications and death also increase with post-term deliveries.
Are term babies actually evolutionary preemies? I haven't read Dr. Karp's book or heard his reasoning, but it seems superficial to me. Medically, from an infant and maternal standpoint, a 38 week delivery is ideal. Over hundreds of thousands of years, the female uterus and pelvis, and the fetal placenta and amniotic fluid found the most favorable time for delivery into a human environment.
If we gestated for two years, like elephants, yes, our babies would be more capable at birth, but we aren't elephants, and mothers and babies would die.
Postterm Pregnancy (Medscape)
Elephant birthing and calves (Wikipedia)