The compromise, in the next round, would effectively do both. In the United Kingdom, where they have an 11-year lead time before the cancellation of the 2021 census, there will be all sorts of interest groups and other considerations in providing data to them.
The compromise, I suggest, is to once again do the long form in its usual fashion--it's presumably prepared--and to simultaneously run a voluntary census with Statistics Canada and the National Statistics Council being responsible for providing a report indicating the degree of bias, where individuals or groups are undercounted, and how that bias is going to affect us in the future.