Ok, we need to get the story straight. If renewable energy sources can provide electricity at a cost that is only slightly higher than what we pay now for cheap fossil supplied oil, and if most people have commutes that are within the range of electric car technology, then what sort of retrofitting is needed?
Because peak oil is not the only reason suburbs need to be retrofitted, just the more urgent one ? Auferre, trucidare, rapere, falsis nominibus imperium; atque, ubi solitudinem faciunt, pacem appellant.
Among other factors, there's one mentioned in the Yglesias posts : elderly population is booming, and driving might not be best for all of them - giving them an option beyond the retirement home could be an idea.
There's the workability of large scale cities : the cars may be electric or gas, but Los Angeles can't grow much more anyway. You can't build enough roads to satisfy an ever increasing traffic.
Also, the land use of suburbia is non-negligible, and one day may need to be used to actually feed people... Auferre, trucidare, rapere, falsis nominibus imperium; atque, ubi solitudinem faciunt, pacem appellant.
And, further, its possible to structure the settlement system so that the same density does not involve as many miles of individual transport.
Of course, the above argument only looks at one dimension of the transition ... it does not involve the ongoing transition we are going to see to a larger share of the population living in more densely populated urban cores, nor what should be done in terms of newly established developments (though with pervasive opportunities for effective infill redevelopment, the pressure for newly established developments is reduced) ... but you increase the average density by increasing the density of existing settlement wherever it lies on the distribution. Utsukushikereba sore de ii
I don't think there's a public transport problem here.
But establishing a regulatory system that mandates that everyone in a suburb lives on 1/4 acre of land that owned, also established consequences in terms of higher drive time (and progressively growing, because sprawl development is a dynamic system that grows traffic) and lower disposable income.
So if some people prefer shorter drive times and higher disposable income, and there is a trade-off between 100% suburban sprawl development and 80% suburban sprawl development where the latter offers the option of shorter drive times and higher disposable income ... why not offer people the choice? Utsukushikereba sore de ii