Granted, at least in the passage above, Smith is looking at a particular time period (immigrants arriving between 1895-99, along with their kids and grandkids) that's different from the present day in several respects. Notably, there was a decent supply of stable and good-paying manufacturing jobs back then—three-fourths of Ford workers in the 1910s, for instance, were immigrants (mostly Eastern European, granted, but I assume Latinos could find similar sorts of jobs)—which can explain why the immigrant families of old could do so well so quickly.

By contrast, economic mobility today is awful, those sorts of manufacturing jobs are hard to come by, and it's reasonable to think that the current generation of Latino immigrants, most of whom arrive quite poor, will have a much harder time ensuring that their kids go to college and get well-paying jobs. But that's because it's harder for everyone on the low end of the income spectrum to do that nowadays, and it's an argument for figuring out ways to improve mobility, not an argument for restricting immigration. There's quite obviously nothing about Latinos per se that makes them "unable" to assimilate.

Now George Borjas, the favorite economist of restrictionists everywhere, has written a paper suggesting that Latinos of old faced "pressures" to assimilate that the current wave of immigrants don't. For instance, immigrants before 1965 were a more diverse lot—you had Latinos and Germans and Italians, etc.—so it was harder for immigrants to stay in their "ethnic enclaves." And there was also, in Borjas' words "an ideological climate that boosted social pressures for assimilation and acculturation" that is no longer around. Well, maybe. Nevertheless, Latino immigrants today seem to be "acculturating" just fine:

A comprehensive 2002 survey of Latinos in the United States by the Pew Hispanic Center and Kaiser Family Foundation provides additional evidence of advancement across generations, particularly in terms of English proficiency. Spanish is the primary language among 72% of first-generation Latinos, but this figure falls to 7% among second-generation Latinos and zero among Latinos who are third generation and higher.

Basically, Latino immigrants and their descendents do very well, especially once one considers that many are exploited and underpaid by their employers, and that they live in a country where economic mobility—not to mention public education—is nothing to brag about. It's also true that if the United States had decided over the past fifty years to help Latin America develop, rather than, you know, fuelling wars, installing various dictatorships, and conducting neoliberal "experiments" on countries like Mexico, then the Latino immigrants who came here would presumably be healthier, wealthier, and better educated and would have "assimilated" more easily. Fun to imagine.