You're going to need a good reference for that one, just because categories have changed, plenty of hispanics have identified as white when they thought it would help and when they could pass, and because plenty of non-citizens don't respond to these types of surveys unless specifically targeted, although I'm sure that there would be methodologies that could attempt to account for them if the investigator intended to.

tl;dr methodology is important when trying to estimate the population of hispanics in the US, especially in 1960. Specific references would be helpful.

Maybe the hispanic people were younger, and had more children (I believe the facts back me up here)? Maybe their religion encouraged this? Maybe hispanic immigrants wanted to go to a nation that already had 3% hispanic population and that was geographically closer to their home nation? Maybe they could trace their family roots over the border?

You appeared to making a relative claim about US and Canadian immigration policy? I thought you were implying that if Canada had an immigration policy like the USA then it would also have 12% black population and that because it doesn't then it's somehow bad and/or cheating?

You'd have to estimate what percentage of African immigrants would have made it into the USA (in the years it wasn't illegal of course) and then substract some kind of modifier similar to the Hispanic immigration, where African immigrants may wish to move somewhere where there already were people who looked like them. And then compare those numbers to see whether Canada was really being exclusionary towards Africans relative to the USA.

how much of that was legal?