I am assuming when you say distributed, you mean (a+b)^3 --> (a^3 + b^3)

This would not be correct - you would need to use cube of binomial rule and that would yield a really big expression.

Normally numerator and denominator in the final result are kept factored (in hope that something from numerator will reduce with something from denominator -- that didn't happen in this problem, but it is the strategy that the solver is using)