Settle on a design for interoperability between different cuda::vector across heterogeneous memory boundaries

[miscco]

We want cuda::vector to be usable accross the board and simplify memory handling for our users.

That requires it to work seamlessly across memory boundaries, e.g how do we envision this to work:

cuda::vector<int, cuda::mr::host_accessible> host_vec { .... };
cuda::vector<int, cuda::mr::device_accessible> device_vec = host_vec;