It’s bad luck for the residents of Tempsford, all 600 of them. Their village in Bedfordhire, in the south-east of England is being eyed as the strategically optimum place to create a new town. New towns were a phenomenon of the postwar years, when successive governments believed that the state should finance the homes needed after the Blitz. Inspired by the utopian Garden City Movement at the turn of the 20^th century, they offered families from soot-blackened terraces in London and elsewhere a new life in supposedly idyllic surroundings. Unfortunately, the architecture of the New Towns didn’t answer the aspiration. The best that can be said of Cumbernauld New Town in Lanarkshire is that the centre, built in Corbusian stilts over a dual carriageway, and some of the housing are now scheduled for demolition.

By the time the last of the new towns, Milton Keynes – not so far from Tempsford – was begun in 1967, the planners had learnt what not to do. Some really good housing was built in the early phases but it was, being publicly funded, anathema to the Thatcher government, which axed the new town corporations. The later parts, built by volume house builders, were rubbish. Besides epic schemes of tree planting, there could be no disguising Milton Keynes’s reliance on the motorcar. It became the city of a thousand roundabouts.

So Tempsford – so temptingly situated between the boom cities of Oxford and Cambridge, so well placed on the rail line between London and Edinburgh – is right to be wary of its fate. This does not mean that, from a national perspective, the building of new towns would be wrong. In theory, they could be a means of providing large numbers of new homes at one go. Once designated, the land could be given its own planning rules to speed up building. Properly conceived, one decently designed new town might be better than ruining a hundred other places. But there’s the rub. How do we know it would be decently designed? Fortunately, one expert source of advice is on hand in the shape of a man, deeply experienced in these matters, whom the prime minister sees every week. This is H.M. the King, whose model development of Poundbury outside Dorchester, the county town of Dorset, now houses 4,600 souls.

There was a time when Poundbury was routinely pooh-poohed by the media, often by journalists who had not been there. But as it nears completion after 30 years, it is easier to see it for what it is: a heroic attempt to create the architectural equivalent of a Morgan sports car, matching the charm of the old with the functionality of the new.

What a battle it was in the beginning. Poundbury flew in the face of the planning orthodoxies of the day. Leon Krier, the master planner, conceived it as a Dorset equivalent of an Italian hill town, whose relatively high urban density would support a good range of shops and services, accessible by foot. Whereas previous housing developments were suburban in character, with gently curving roads along which motorists could glide unobstructed, the streets of Poundbury were made deliberately confusing to navigate. Traffic can only move slowly, giving pedestrians the upper hand. Walking not only provides exercise but encourages neighbourly contact; there are pleasant places in which to chat. The zoning rules were thrown out of the window, so that workplaces could be scattered among residential accommodation, meaning that (in theory) inhabitants did not need to commute. Critics, who did not always make the effort to visit the place, sneered at the ‘toy town’ character of the streets, which were broadly vernacular in character with some classical flourishes.

Perhaps too many architects were involved at times, causing the pudding to be over-egged. Better too much architecture, though, than none, as is the case with the work of the volume house builders. Over the decades, the local materials of which Poundbury is built have weathered attractively; the trees lining the streets have grown up; there is a school, and some impressive play facilities for children have been created. God is in the detail, said the Modernist architect Mies van der Rohe: at Poundbury that means carefully siting lampposts so that they don’t obstruct the pavement, or – even better – attaching them to nearby buildings so they don’t touch the pavement at all.

It is sometimes assumed that Poundbury is so attractive that it must be middle-class. Instead, a third of the accommodation is affordable. You would not necessarily know it was there, because rather than being huddled into a ghetto on the worst part of the site, it is dotted throughout; there is nothing visually to differentiate it from the homes sold at market rates. These principles have been developed at the Duchy of Cornwall’s other major project, outside Newquay in Cornwall. Newquay is not in a prosperous part of the country but Nansledan is thriving; there are young families out and about, and a new high street full of flourishing shops – a signal contrast to most high streets in Britain. There are fewer architectural setpieces than Poundbury, but a greater emphasis on sustainability: this goes beyond espaliered fruit trees and bat boxes to a policy to support local granite quarries, craftspeople and food producers. Doesn’t this provide the perfect model for Labour to follow in its attempt to boost house building? Nansledan is popular with the folk who live there. Poundbury not only provides homes but work for 2,400 people. What’s not to like?

There are several flies in the ointment, however. For one thing, neither Poundbury nor Nansledan is a new town: they are extensions to existing towns. While they are exceptionally well supplied with services at a local level, they rely on Dorchester and Newquay for rail links, hospitals and the like. Economically they make sense because the Duchy of Cornwall already owned the land on which they’re built; having been in existence since the 14th century, it could afford to take a long view, accepting that upfront investment in infrastructure would cost money at the beginning of the project, in the knowledge that land prices would rise and yield a greater profit by the end.

The government could take a similar approach at Tempsford, if it bought the site through compulsory purchase – but of course it won’t; the cost would create one of those famous black holes in the budget. As for providing a significant contribution to the target of 1.5 million new homes before the next election, forget it. Whatever Angela Rayner, the housing minister, does to liberalise the planning system, it is unlikely that a new town could get planning permission in under five years – 15 years is more likely under the current rules.

It takes decades to complete a Poundbury or Nansledan. Fewer than 100 properties are released onto the market each year. Any more would risk flooding the market so that the house builders would not get the return needed to keep them in business. This is true not only for Poundbury but all the other major developments it has inspired, such as Tornagrain outside Inverness in the Scottish Highlands.

Think of what’s needed. As an architect observed to me the other day, there are about 1,250 working days before the next election. That means that the Labour government will have to build 1,200 a day – that’s five Poundburies every month. Where are the bricklayers? Where are the bricks? New towns may be a good idea if well planned – but there’s no point in rushing when they can’t meet the objective the government has set itself. Better to stay calm, go at a slower pace, while reviewing other options for quick delivery – for example, all those empty office buildings that could easily become flats.