> There is a well-known proof from information theory [...] Remember that this proof is talking in terms of the number of comparisons necessary to sort n items

> There is a well-known proof from information theory [...] Remember that this proof is talking in terms of the number of comparisons necessary to sort n items. The moment you stop comparing data, like in radix sort (or more generally, bucket sort), that all flies out the window.