"It took 20 mins from "uh ho this is bad" to baby in DHs arms and even with the fastest transfer that would have been at least doubled if I had been at home."
You are absolutely right that some babies will be better off born in hospital. And we can't predict which babies these will be. Some women will be unlucky and have a cord prolapse/placental abruption at home and those women will probably have worse outcomes than had they been in hospital.
HOWEVER, the fact that the neonatal outcomes for everyone except first time mums are comparable means that some babies must be having their well-being compromised by being born in hospital. It must be the case, otherwise you'd see higher deaths at homebirths than in hospital births.
Interventions in labour (which are more likely to happen in hospital) are helpful to a small number of babies. For the rest of the mothers and babies who receive them they disrupt labour and put the health of both at risk. Luckily for those mothers though, there are doctors at hand to 'fix' the problems that giving birth in hospital causes to labouring women and babies.
FWIW - it's very sobering to look at the c-section and maternal death rates from the 1950's, because it tells a very different story than the one people imagine. The c/s rate in the 1950's was 3%. And the maternal death rate was slightly less than one in a thousand. And that includes all the deaths from botched abortions and thrombo-embolisms (probably more common in the days when women were made to have 10 days complete bed rest following birth).
For me this suggests that the VAST majority of women now who are having sections for 'failure to progress' would have had a vaginal birth back in the 1950's. In other words, very few babies are truly 'undeliverable' vaginally, even in a population which probably had higher rates of skeletal abnormalities (many women delivering in the 1950's would have been born in the 1920's and 1930's, when childhood malnutrition was much, much more common than today).
The neonatal mortality rate has gone down from about 2 or 3 in a hundred to 1 in 200 in that time, which is a very big fall, but it's hard to work out how much of that is due to advances in diagnostic techniques, so that far fewer babies now are born with congenital abnormalities with a very poor prognosis.