I think it depends on the outcomes you want at the end. If the outcome you want is for Canada to generate new knowledge and keep us at the top of the pointy bit of science, then definitely universities are where you go. I sometimes say that colleges will not discover the Higgs boson or detect gravitational waves. That's been done already, although I may have to change my analogies. If what you want is economic development, jobs and small companies getting bigger in regions, then this is a type of impact we can have.
From the point of view of whether it is better to have college programs or have college and universities applying for the same programs, I think you should evaluate if a proposal is good based on the outcomes you want out of the research. If what you want is something that universities are excellent at bringing, then I think it's a university program. If it's something that colleges are excellent at bringing, then make it a college program.
There could be some that have both that apply, but then at that point, the evaluation criteria and the outcomes you want and expect should be adapted such that colleges won't be evaluated on the number of papers they write in nature or science, because this is not what we do. This is not our output. Our outputs are prototypes. They are new products that get commercialized so that our companies do better.