Now, actually I've come to the conclusion that the dominance of Toronto's immigrant languages has been quite overstated, by North American standards. This is from realizing that though about a third of GTA residents speak a non-official language at home (as of the last stats, since the latest census won't have this data until later in the year), many US cities actually rank higher in terms of non-English speakers.
Half of New Yorkers and Angelenos speak a non-English language at home, according to the American Community Survey. Much of California's cities are in the 40-50% range in terms of people who speak an immigrant language at home. 30% of Chicagoans do, which is closer to Toronto's figure. Yet all these places have a lower foreign born percentage than Toronto does.
Miami, which is half foreign born and the only US city with a higher immigrant proportion than Toronto, has nearly three quarters of its citizens speaking a non-English language at home (I'd imagine it's mostly Spanish).
This is surprising; in these big immigrant-rich US cities, the share of those speaking an immigrant language at home is higher than the share of immigrants.
But in Toronto, the share of those speaking an immigrant language is actually lower than the share of immigrants..
Perhaps more of Toronto's immigrants come already knowing and regularly speaking with family in English (eg. Caribbean immigrants). Or alternatively perhaps, Toronto, contrary to popular belief, is actually assimilating its immigrants to speaking English more than stateside, or big urban US cities are actually hanging on to their immigrant languages more. But that's not what the classic narrative we all hear about the melting pot and mosaic is. Perhaps one major difference is that stateside, immigrant language retention is more driven by one major language, Spanish, while for Canada, in a city like Toronto, you don't have one immigrant language in particular dominating as much -- maybe a larger number of diverse immigrant languages with smaller shares of speakers are lost more easily than say one immigrant language like Spanish with a larger share of a given city's speakers (but maybe not, as NYC has retained linguistic diversity for so long).
Granted, I have no idea how comparable the US and Canadian data is -- perhaps the amount of speaking that one does to consider answering the survey question might be different. But it goes to show that Toronto isn't as different or exceptional as we often make it out to be.