You're mixing up two things here. It doesn't really matter if the recommendation algorithm is good or not.
The advancement isn't that we now have a better recommendation algorithm, it's that we thought that there was a problem that was impossible to solve with current computers and was being used as example of a problem that only quantum computing could solve and we've now learned that that isn't the case.
The advancement isn't that we now have a better recommendation algorithm, it's that we thought that there was a problem that was impossible to solve with current computers and was being used as example of a problem that only quantum computing could solve and we've now learned that that isn't the case.