Disallow substitution sensitivity where N is unknown

[Shoeboxam]

We should eliminate the class of queries on unsized datasets where the definition of neighboring is based on the substitution distance. The set of neighboring datasets excludes all datasets with a different size, as the substitution distance between any two datasets of a different size is infinite. This is a pretty ill construction, as it leads to strange cases where we artificially inflate the sensitivity- like the sensitivity of count queries in the substitution distance.