That doesn't help at all because although Datalog is a syntactic fragment, its dynamic semantics are very different, so the operations defined in Prolog standard are suggestive at best and completely nonsensical at worst.
For example, the program `p(X) :- p(X)` with query `?- p(X)` is an infinite loop in Prolog but in Datalog it terminates trivially and is equal to the empty set of facts for `p` in the absence any other facts.
Also people don't like to switch and SQL does a good enough job for most cases and recently it added a lot of fancy features like windowing functions.