I'm very skeptical about the applicability of group theory to computer science - outside of the usual applications to counting/combinatorics and crypto/number theory.