It depends on the type of people who get into office, I mean despite FDR's prestige in the present day, it was still rather touch and go during his time in the 1930's. If he hadn't tried to roll back the New Deal after 1937, unemployment numbers might've continued dropping down, reaching maybe 10% by 1940. And abroad, the rise of Stalin in the Soviet Union, Militarism in Japan, and Nazism in Germany were hardly a sure thing in the 1920's, and it wouldn't be especially difficult to butterfly all those in a manner that leaves the 1940's peaceful.
You have to butterfly away the Treaty of Versailles to get rid of Hitler or somebody like him, and Japan's militarism was a consequence of their demand for raw materials and the fact that they were not keen on being dictated to by other nations. Stalin can be butterflied away, but the other two are rather hard, and while getting rid of Hitler isn't hard, that won't stop the problems that Versailles added to Germany and the resentment it caused.
I've never said anything about stopping suburbanization, just slowing it down and changing it's character without the billions of dollars in subsidies and capital that caused the boom OTL.
In favor of what, though? America's attempts at building high-rises to provide modern housing in their cities were at the very least not sky-high successes (In New York, for example) at worst, with places like Pruitt-Igoe in St. Louis and the Cabrini-Green and Robert Taylor Homes projects of Chicago, outright disasters that turned rough neighborhoods into outright shitholes. Yes, more money poured into these areas will reduce problems, but especially with desegregation looming, there are gonna be problems in these cities. Population growth will force expansions, but even if you try to have more people live in the cities, what happened at Pruitt-Igoe and Cabrini-Green and the Robert Taylor Homes is gonna cause big turnoffs by the early to mid 1960s, and more people will want single family homes, which means sprawl begins all over again.
From a social and civil engineering perspective, higher-density neighborhoods have many advantages, but trying to get that in post-war America is extremely hard, as you'll find the government battling the wants of the people and many developers, as well as builders like Moses. Beating all of that would require vast mass transit systems making it possible for people to get where they want to go quickly and easily and at any time they wish, which is expensive and probably not feasible, even if you gave many of the transit companies enormous power in the post-war era.
A better idea might be for the civil rights movement to happen but with many people choosing to not abandon the neighborhoods they've lived in all their lives, and this causing comfortably integrated neighborhoods by the early 1970s, followed by an energy crisis that lasts longer and has a harder hit on the American psyche, to the point that the car-oriented lifestyle of the suburbs seems less desirable than the denser neighborhoods of the cities. The broken windows theory comes fifteen years earlier, and is widely used in several American cities. While projects like Pruitt-Igoe and the Chicago projects are still messes, the combination of broken-windows policing strategies and active residents leads to major cleanups of neighborhoods in many American cities. Combined with what I said earlier about keeping America's middle class strong, and you can by the 1980s see the movement into the suburbs reversed, with people wanting to raise children in the more walkable, cosmopolitan neighborhoods of the cities themselves, causing large sections of old suburbs to be torn down or redeveloped in the 1990s and 2000s.
BULLSHIT. And I'll say it again BULLSHIT. It's all about the amount of support cities were willing to give, which was very little until the past 40 years.
First, Los Angeles is actually a very dense city, it just has a shitload of parkland and low-density areas within city limits than deflates it's density somewhat (Several of the cities neighboring LA are better indicators of LA's density than LA's average, the metro area as a whole is also surprisingly dense). Secondly, while it's geography is inferior to cities like San Francisco and New York in favorability to transit, it does a fair number of geographic "pinch points" that give boosts to Public Transit, and it's Downtown has the fortune of sitting of sitting at the exact center of the metro area, with nearly all major corridors feeding into it. Thirdly, LA had a tremendous mass transit system in the form of the Red and Yellow cars that could've very survived had they not been brought under control of people who sought to dismantle them, and with them, LA could've easily had transit ridership several times that of OTL.
All of that is very true (about the denser portions, that is - most of the Sunset, Santa Monica and Wilshire Boulevard regions are quite dense), but Los Angeles is still a very, very spread out city, you can drive for hours there and not leave the city there, and as a result its car usage is still going to end being very high, even if you do massively increase its mass transit usage. (I lived there for four years, man, and I took the subways lots of times, but the city is not mass transit friendly.) Even if you kept the Pacific Electric system and built subways and heavy rail to complement them, you will still never get to 25% of trips done by mass transit. You might get 10-15%, which would still make for a public service that very easily covers its costs or a quite profitable private business, but you will still have a pile of cars on the road in Los Angeles. Population growth in Los Angeles also works against the anti-sprawl idea - LA's population more than doubled between 1920 and 1930, and it kept on going after that - the city itself grew from 577,000 in 1920 to 1,970,000 in 1950 to 2,967,000 in 1980, and that's just the city of Los Angeles, not counting the explosive growth in the cities between Long Beach and the city center (Torrance, Compton, Inglewood, Marina del Rey, Redondo Beach, Lakewood) and the cities around the city of Los Angeles itself (Pasadena, Glendale, Santa Monica, Santa Clarita, Downey) and the inland empire communities (Ontario, Riverside, Pomona, West Covina) or Orange County. The idea packing all of these closer to downtown LA goes against the very idea of what California was to many of those who went there during the 20th Century. Yes, you can get the PE to serve most of these places, and you could get Southern Pacific Lines and the Atchison, Topeka and Santa Fe to provide commuter rail service to most of these areas, but that won't stop the car culture and it won't stop the suburban growth.
Secondly, Detroit and Chicago were very dense cities that also had well developed streetcars systems, and Chicago's streetcar system actually extended into the suburbs, and was one of the largest in the world, carrying more than three times as many passengers as the L did (about 1.9 million to the metro's 660,000) , with even more on bus and rail. If that system was modernized properly (connecting the Commuter lines and Streetcar lines with subways under the central area), then getting ridership up to proportion I suggested would be simple.
To 25% of trips with how wide the Chicago area is? Not gonna happen, man. I don't mean to be a downer, but you aren't gonna get everyone who lives in Gary and Joliet and Hammond and Waukegan who travel to Chicago to work onto trains every time. Likewise, Chicago's center is dense, but its subrubs are quite wide as well. Chicago and 2012 Toronto have a lot of similarities, and while both cities have excellent (and well-patronized) commuter rail systems and mass transit systems (which certainly could be improved further), you aren't gonna get 25% of all trips in either place to go by transit instead of cars without raising fuel prices a lot, and as I said before, the odds of that happening in America are very slim.
Houston is a special case, as they were one of a few major cities that abandoned their streetcar systems before WWII. And with the type of POD I'm mentioning, it's possible that Miami might not be a major city TTL, even so Miami too actually has favorable geography for Mass Transit, as like Salt Lake City, the Urbanized area is a narrow strip of land running North to South. And Atlanta might not even be as big as it is if wasn't for WWII, which caused it to turn into a major industrial hub.
Miami was a pretty remote city in the greater scheme of things until Castro caused a mountain of Cubans to flee to South Florida and act as a catalyst for the development of the whole area, so you could make that one into a transit-heavy city, but that has a political problem because of said Cubans, who will not be keen on government efforts to make the city center denser - and after about 1970, those guys dominated the city's politics to such a degree that Miami might be the only city in the Western Hemisphere to not accept a visit by Nelson Mandela - because he thanked Castro for his support of the ANC during its time as a banned organization in South Africa. Salt Lake City is pretty similar to Miami, except it was much slower to grow - the Salt Lake City metro area didn't pass one million people until 2001, and the fact that the metro area of Salt Lake is divided rather heavily and local governments are not particularly in favor of government-support for transportation schemes doesn't help matters there.
Atlanta's development has many things going for it which stopping WWII isn't gonna stop, namely its status as a major transport hub. And like Los Angeles, you could make its mass transit far better, but that won't stop the growth of car culture.
FDW, I know where you are going with this, but keeping while said interurbans and mass transit networks going during the post-WWII era is possible, it's not gonna knock off a huge number of trips done by cars. The best you can hope for outside of very dense cities like New York is about 15% of trips being done using mass transit. That's still billions of trips in a year in American cities, mind you, but its not gonna stop Detroit rise and its not gonna stop the introduction of many cars, and population growth is invariably going to make for more cars out there as well - the population of the United States rose 106.0 million in 1920 to 151.3 million in 1950 to 226.5 million in 1980. The cities are gonna grow out, particularly as the growth of both manufacturing and service industries grows major cities. You can't avoid this short of measures that are simply ASB for the post-war United States.
And even if you were to get 20% of the American car market cut away because of mass transit, you're still looking at a car market of millions. The record for most cars registered in a year in the US (in 1985) is 11.1 million, and the lowest number in modern times was in 2009, with a number of new cars sold in the United States was 5.6 million. Taking away 20% of either of those numbers still leaves a vast sum, more than enough to still make GM, Ford and Chrysler into industrial giants. Short of busting the four US automakers into six or seven, you won't stop that growth.