I'm glad that Monsieur Fortin provided those data. I know that, as Professor Gingras said, the tri-councils follow these data very closely. To the best of my knowledge, I've not seen that there's a big discrepancy. The success rates do not seem to be heavily biased by the language of submission. I could be wrong, because this is not something I looked at yesterday.
In terms of the environment, if you're a scientist, you think very carefully about what language you're going to submit your application in. If you are going to try to write it in English and you're francophone, you're going to have someone help you and vice versa. That's what I tend to see.
I have to say that McGill in my experience is much more English. I don't really consider McGill to be bilingual, per se, but I think most people or many of the applicants from McGill are functional in both French and English.